rowid,run.accession,experiment.accession,sample.accession,study.accession,bioproject,study.title,study.alias,study.type,study.abstract,study.attributes,study.PMIDs,sample.description,sample.title,sample.alias,sample.centername,sample.attributes,GEOsample.title,GEOsample.dataprocessing,GEOsample.source,GEOsample.treatmentprotocol,GEOsample.extractprotocol,GEOsample.growthprotocol,GEOsample.characteristics,GEOsample.accession,experiment.title,experiment.alias,experiment.library_name,experiment.design_description,experiment.library_construction_protocol,experiment.attributes,experiment.library_strategy,experiment.library_source,experiment.library_selection,experiment.library_layout,experiment.platform,experiment.instrument_model,experiment.spot_descriptor,experiment.study_ref,run.title,run.attributes,run.filename,run.semantic_name,run.total_bases,run.total_spots,run.alias,run.read_lengths,run.base_counts,run.r1_length,run.r2_length,run.r3_length,run.r4_length,run.Acount,run.Ccount,run.Gcount,run.Tcount,run.Ncount,run.experiment,run.pool_member,submission.accession,submission.srasource,submission.bioprojectsource,seqdetective.n_mates,seqdetective.mapping_rate.mate1,seqdetective.mapping_rate.mate2,seqdetective.nofeature_rate.mate1,seqdetective.nofeature_rate.mate2,seqdetective.sparsity.mate1,seqdetective.sparsity.mate2,seqdetective.pos_strand_rate.mate1,seqdetective.pos_strand_rate.mate2,seqdetective.readlen.mate1,seqdetective.readlen.mate2,seqdetective.judgement.mate1,seqdetective.judgement.mate2,seqdetective.judgement.reason,platform_family,instrument_generation,read_bias,selection_class,prep_kit,sc_or_bulk,tech_class,technology,tech_variant,submission.bioprojectsource.country,earliest_date,devstage_curation,devstage_curation_coarse,tissue_curation,tissue_curation_coarse
29202,SRR27329257,SRX23006269,SRS19970080,SRP479683,PRJNA1055904,BMP dependent patterning of ectoderm tissue material properties modulates lateral mesendoderm cell migration during early zebrafish gastrulation,GSE251904,Transcriptome Analysis,Cell migration is a fundamental process during embryonic development. Most studies in vivo have focussed on the migration of cells using the extracellular matrix ECM as their substrate for migration. In contrast  much less is known about how cells migrate on other cells  as found in early embryos when the ECM has not yet formed. Here  we show that lateral mesendoderm LME cells in the early zebrafish gastrula use the ectoderm as their substrate for migration. We show that the lateral ectoderm is permissive for the animal pole directed migration of LME cells  while the ectoderm at the animal pole halts it. These differences in the permissiveness depend on the lateral ectoderm being more cohesive than the animal ectoderm  a property controlled by BMP signalling within the ectoderm. Collectively  these findings identify ectoderm tissue cohesion as one critical factor regulating LME migration during zebrafish gastrulation. Overall design: To investigate the differences in gene expression between animal and lateral ectoderm causing the difference in material properties between the two ectodermal tissues.,,pubmed:40057955,,Zebrafish Lateral Ectoderm 7.5 hpf 4,GSM7989708,,source name:Lateral Ectoderm|tissue:Lateral Ectoderm|genotype:WT TgSebox::eGFP embryos|treatment:1 cell stage injection of PAmCherry1 mRNA  Photoactivation  FACS|geo loc name:missing|collection date:missing,Zebrafish Lateral Ectoderm 7.5 hpf 4,FastQC quality control Trimmomatic adapter & quality trimming; parameters: 2:30:10 LEADING:3 TRAILING:3 SLIDINGWINDOW:4:15 MINLEN:19 FastQC controlling trimming results salmon Alignment; used flags:   seqBias   qcBias R   DESeq2 differential gene expression analysis; DESEq2  apeglm  tximport  genefilter  GenomicRanges R   EnhancedVolcano & biomaRt visualization Assembly: Danio rerio GRCz11 Supplementary files format and content: tab delimited text file with gene name  gene length  effective length  tpm and number of reads for each sample,Lateral Ectoderm,Embryo were injected with 125 pg of PAmCherry1 mRNA at xxx cell stage. At 6 hpf  embryos were mounted for up right imaging and a 444 x 220 x 58.5 µm volume was photoactivated using a 405 nm laser either at the animal pole or on the lateral side. The photoactivation process was composed of 10 cycles of 28 sec per embryo in a time window of approximately 90 min. post photoactivation  embryos were transferred in Ca2+ free Ringer solution and dissociated by gently pipetting using a P 1000 pipette to obtain a single cell suspension. 5000 PAmCherry1+ cells were sorted in 150 µl of RTL plus lysis buffer RNeasy Plus Micro Kit  QIAGEN with 1% of 2 mercaptoethanol.,Cells were subjected to RNA extraction using RNeasy Plus Micro Kit QIAGEN according to the manufacturer instructions Complete cDNA synthesis and library preparation were performed using SMART Seq v3 protocol with Nextera UDI adapters. Libraries were then quantified by qPCR KAPA Biosysytems.,Embryos were kept at 28.5 C until dissociation,tissue:Lateral Ectoderm|genotype:WT TgSebox::eGFP embryos|treatment:1 cell stage injection of PAmCherry1 mRNA  Photoactivation  FACS,GSM7989708,GSM7989708: Zebrafish Lateral Ectoderm 7.5 hpf 4; Danio rerio; RNA Seq,GSM7989708 r1,GSM7989708,1,Cells were subjected to RNA extraction using RNeasy Plus Micro Kit QIAGEN according to the manufacturer instructions Complete cDNA synthesis and library preparation were performed using SMART Seq v3 protocol with Nextera UDI adapters. Libraries were then quantified by qPCR KAPA Biosysytems.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina NovaSeq 6000,,SRP479683,,,Lateral_Ectoderm_4.mate1.fastq.gz Lateral_Ectoderm_4.mate2.fastq.gz,fastq fastq,6211000588.0,26929501.0,GSM7989708 r1,0:80.64 1:150,A:1950010947;C:1101223011;G:1497556265;T:1656016608;N:6193757,80,150,,,1950010947,1101223011,1497556265,1656016608,6193757,SRX23006269,SRS19970080,SRA1774740,"Heisenberg group, Institute of Science and Technology Austria (ISTA)","Heisenberg group, Institute of Science and Technology Austria (ISTA)",2,0.65918,0.4633,0.04875,0.02201,0.90623,0.92608,0.79936,0.77887,68,150,B,B,mate2-mate1 similar by mapping diff,illumina,novaseq_era,unknown,cdna_unspecified,nextera,sc,single_cell_plate,smartseq,,Austria,2023-12-22,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
29203,SRR27329258,SRX23006268,SRS19970081,SRP479683,PRJNA1055904,BMP dependent patterning of ectoderm tissue material properties modulates lateral mesendoderm cell migration during early zebrafish gastrulation,GSE251904,Transcriptome Analysis,Cell migration is a fundamental process during embryonic development. Most studies in vivo have focussed on the migration of cells using the extracellular matrix ECM as their substrate for migration. In contrast  much less is known about how cells migrate on other cells  as found in early embryos when the ECM has not yet formed. Here  we show that lateral mesendoderm LME cells in the early zebrafish gastrula use the ectoderm as their substrate for migration. We show that the lateral ectoderm is permissive for the animal pole directed migration of LME cells  while the ectoderm at the animal pole halts it. These differences in the permissiveness depend on the lateral ectoderm being more cohesive than the animal ectoderm  a property controlled by BMP signalling within the ectoderm. Collectively  these findings identify ectoderm tissue cohesion as one critical factor regulating LME migration during zebrafish gastrulation. Overall design: To investigate the differences in gene expression between animal and lateral ectoderm causing the difference in material properties between the two ectodermal tissues.,,pubmed:40057955,,Zebrafish Lateral Ectoderm 7.5 hpf 3,GSM7989707,,source name:Lateral Ectoderm|tissue:Lateral Ectoderm|genotype:WT TgSebox::eGFP embryos|treatment:1 cell stage injection of PAmCherry1 mRNA  Photoactivation  FACS|geo loc name:missing|collection date:missing,Zebrafish Lateral Ectoderm 7.5 hpf 3,FastQC quality control Trimmomatic adapter & quality trimming; parameters: 2:30:10 LEADING:3 TRAILING:3 SLIDINGWINDOW:4:15 MINLEN:19 FastQC controlling trimming results salmon Alignment; used flags:   seqBias   qcBias R   DESeq2 differential gene expression analysis; DESEq2  apeglm  tximport  genefilter  GenomicRanges R   EnhancedVolcano & biomaRt visualization Assembly: Danio rerio GRCz11 Supplementary files format and content: tab delimited text file with gene name  gene length  effective length  tpm and number of reads for each sample,Lateral Ectoderm,Embryo were injected with 125 pg of PAmCherry1 mRNA at xxx cell stage. At 6 hpf  embryos were mounted for up right imaging and a 444 x 220 x 58.5 µm volume was photoactivated using a 405 nm laser either at the animal pole or on the lateral side. The photoactivation process was composed of 10 cycles of 28 sec per embryo in a time window of approximately 90 min. post photoactivation  embryos were transferred in Ca2+ free Ringer solution and dissociated by gently pipetting using a P 1000 pipette to obtain a single cell suspension. 5000 PAmCherry1+ cells were sorted in 150 µl of RTL plus lysis buffer RNeasy Plus Micro Kit  QIAGEN with 1% of 2 mercaptoethanol.,Cells were subjected to RNA extraction using RNeasy Plus Micro Kit QIAGEN according to the manufacturer instructions Complete cDNA synthesis and library preparation were performed using SMART Seq v3 protocol with Nextera UDI adapters. Libraries were then quantified by qPCR KAPA Biosysytems.,Embryos were kept at 28.5 C until dissociation,tissue:Lateral Ectoderm|genotype:WT TgSebox::eGFP embryos|treatment:1 cell stage injection of PAmCherry1 mRNA  Photoactivation  FACS,GSM7989707,GSM7989707: Zebrafish Lateral Ectoderm 7.5 hpf 3; Danio rerio; RNA Seq,GSM7989707 r1,GSM7989707,1,Cells were subjected to RNA extraction using RNeasy Plus Micro Kit QIAGEN according to the manufacturer instructions Complete cDNA synthesis and library preparation were performed using SMART Seq v3 protocol with Nextera UDI adapters. Libraries were then quantified by qPCR KAPA Biosysytems.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina NovaSeq 6000,,SRP479683,,,Lateral_Ectoderm_3.mate1.fastq.gz Lateral_Ectoderm_3.mate2.fastq.gz,fastq fastq,5066130338.0,22014930.0,GSM7989707 r1,0:80.12 1:150,A:1476750755;C:934859886;G:1246347548;T:1403280136;N:4892013,80,150,,,1476750755,934859886,1246347548,1403280136,4892013,SRX23006268,SRS19970081,SRA1774740,"Heisenberg group, Institute of Science and Technology Austria (ISTA)","Heisenberg group, Institute of Science and Technology Austria (ISTA)",2,0.7566,0.52272,0.0462,0.02684,0.89489,0.91246,0.78848,0.77633,68,150,B,B,mate2-mate1 similar by mapping diff,illumina,novaseq_era,unknown,cdna_unspecified,nextera,sc,single_cell_plate,smartseq,,Austria,2023-12-22,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
29204,SRR27329259,SRX23006267,SRS19970078,SRP479683,PRJNA1055904,BMP dependent patterning of ectoderm tissue material properties modulates lateral mesendoderm cell migration during early zebrafish gastrulation,GSE251904,Transcriptome Analysis,Cell migration is a fundamental process during embryonic development. Most studies in vivo have focussed on the migration of cells using the extracellular matrix ECM as their substrate for migration. In contrast  much less is known about how cells migrate on other cells  as found in early embryos when the ECM has not yet formed. Here  we show that lateral mesendoderm LME cells in the early zebrafish gastrula use the ectoderm as their substrate for migration. We show that the lateral ectoderm is permissive for the animal pole directed migration of LME cells  while the ectoderm at the animal pole halts it. These differences in the permissiveness depend on the lateral ectoderm being more cohesive than the animal ectoderm  a property controlled by BMP signalling within the ectoderm. Collectively  these findings identify ectoderm tissue cohesion as one critical factor regulating LME migration during zebrafish gastrulation. Overall design: To investigate the differences in gene expression between animal and lateral ectoderm causing the difference in material properties between the two ectodermal tissues.,,pubmed:40057955,,Zebrafish Lateral Ectoderm 7.5 hpf 2,GSM7989706,,source name:Lateral Ectoderm|tissue:Lateral Ectoderm|genotype:WT TgSebox::eGFP embryos|treatment:1 cell stage injection of PAmCherry1 mRNA  Photoactivation  FACS|geo loc name:missing|collection date:missing,Zebrafish Lateral Ectoderm 7.5 hpf 2,FastQC quality control Trimmomatic adapter & quality trimming; parameters: 2:30:10 LEADING:3 TRAILING:3 SLIDINGWINDOW:4:15 MINLEN:19 FastQC controlling trimming results salmon Alignment; used flags:   seqBias   qcBias R   DESeq2 differential gene expression analysis; DESEq2  apeglm  tximport  genefilter  GenomicRanges R   EnhancedVolcano & biomaRt visualization Assembly: Danio rerio GRCz11 Supplementary files format and content: tab delimited text file with gene name  gene length  effective length  tpm and number of reads for each sample,Lateral Ectoderm,Embryo were injected with 125 pg of PAmCherry1 mRNA at xxx cell stage. At 6 hpf  embryos were mounted for up right imaging and a 444 x 220 x 58.5 µm volume was photoactivated using a 405 nm laser either at the animal pole or on the lateral side. The photoactivation process was composed of 10 cycles of 28 sec per embryo in a time window of approximately 90 min. post photoactivation  embryos were transferred in Ca2+ free Ringer solution and dissociated by gently pipetting using a P 1000 pipette to obtain a single cell suspension. 5000 PAmCherry1+ cells were sorted in 150 µl of RTL plus lysis buffer RNeasy Plus Micro Kit  QIAGEN with 1% of 2 mercaptoethanol.,Cells were subjected to RNA extraction using RNeasy Plus Micro Kit QIAGEN according to the manufacturer instructions Complete cDNA synthesis and library preparation were performed using SMART Seq v3 protocol with Nextera UDI adapters. Libraries were then quantified by qPCR KAPA Biosysytems.,Embryos were kept at 28.5 C until dissociation,tissue:Lateral Ectoderm|genotype:WT TgSebox::eGFP embryos|treatment:1 cell stage injection of PAmCherry1 mRNA  Photoactivation  FACS,GSM7989706,GSM7989706: Zebrafish Lateral Ectoderm 7.5 hpf 2; Danio rerio; RNA Seq,GSM7989706 r1,GSM7989706,1,Cells were subjected to RNA extraction using RNeasy Plus Micro Kit QIAGEN according to the manufacturer instructions Complete cDNA synthesis and library preparation were performed using SMART Seq v3 protocol with Nextera UDI adapters. Libraries were then quantified by qPCR KAPA Biosysytems.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina NovaSeq 6000,,SRP479683,,,Lateral_Ectoderm_2.mate1.fastq.gz Lateral_Ectoderm_2.mate2.fastq.gz,fastq fastq,6372234032.0,27787436.0,GSM7989706 r1,0:79.32 1:150,A:1908123129;C:1195826989;G:1555786583;T:1706205499;N:6291832,79,150,,,1908123129,1195826989,1555786583,1706205499,6291832,SRX23006267,SRS19970078,SRA1774740,"Heisenberg group, Institute of Science and Technology Austria (ISTA)","Heisenberg group, Institute of Science and Technology Austria (ISTA)",2,0.52417,0.50653,0.02922,0.02493,0.89479,0.90938,0.74721,0.74563,68,150,B,B,biological fallback assumption,illumina,novaseq_era,unknown,cdna_unspecified,nextera,sc,single_cell_plate,smartseq,,Austria,2023-12-22,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
29205,SRR27329260,SRX23006266,SRS19970079,SRP479683,PRJNA1055904,BMP dependent patterning of ectoderm tissue material properties modulates lateral mesendoderm cell migration during early zebrafish gastrulation,GSE251904,Transcriptome Analysis,Cell migration is a fundamental process during embryonic development. Most studies in vivo have focussed on the migration of cells using the extracellular matrix ECM as their substrate for migration. In contrast  much less is known about how cells migrate on other cells  as found in early embryos when the ECM has not yet formed. Here  we show that lateral mesendoderm LME cells in the early zebrafish gastrula use the ectoderm as their substrate for migration. We show that the lateral ectoderm is permissive for the animal pole directed migration of LME cells  while the ectoderm at the animal pole halts it. These differences in the permissiveness depend on the lateral ectoderm being more cohesive than the animal ectoderm  a property controlled by BMP signalling within the ectoderm. Collectively  these findings identify ectoderm tissue cohesion as one critical factor regulating LME migration during zebrafish gastrulation. Overall design: To investigate the differences in gene expression between animal and lateral ectoderm causing the difference in material properties between the two ectodermal tissues.,,pubmed:40057955,,Zebrafish Lateral Ectoderm 7.5 hpf 1,GSM7989705,,source name:Lateral Ectoderm|tissue:Lateral Ectoderm|genotype:WT TgSebox::eGFP embryos|treatment:1 cell stage injection of PAmCherry1 mRNA  Photoactivation  FACS|geo loc name:missing|collection date:missing,Zebrafish Lateral Ectoderm 7.5 hpf 1,FastQC quality control Trimmomatic adapter & quality trimming; parameters: 2:30:10 LEADING:3 TRAILING:3 SLIDINGWINDOW:4:15 MINLEN:19 FastQC controlling trimming results salmon Alignment; used flags:   seqBias   qcBias R   DESeq2 differential gene expression analysis; DESEq2  apeglm  tximport  genefilter  GenomicRanges R   EnhancedVolcano & biomaRt visualization Assembly: Danio rerio GRCz11 Supplementary files format and content: tab delimited text file with gene name  gene length  effective length  tpm and number of reads for each sample,Lateral Ectoderm,Embryo were injected with 125 pg of PAmCherry1 mRNA at xxx cell stage. At 6 hpf  embryos were mounted for up right imaging and a 444 x 220 x 58.5 µm volume was photoactivated using a 405 nm laser either at the animal pole or on the lateral side. The photoactivation process was composed of 10 cycles of 28 sec per embryo in a time window of approximately 90 min. post photoactivation  embryos were transferred in Ca2+ free Ringer solution and dissociated by gently pipetting using a P 1000 pipette to obtain a single cell suspension. 5000 PAmCherry1+ cells were sorted in 150 µl of RTL plus lysis buffer RNeasy Plus Micro Kit  QIAGEN with 1% of 2 mercaptoethanol.,Cells were subjected to RNA extraction using RNeasy Plus Micro Kit QIAGEN according to the manufacturer instructions Complete cDNA synthesis and library preparation were performed using SMART Seq v3 protocol with Nextera UDI adapters. Libraries were then quantified by qPCR KAPA Biosysytems.,Embryos were kept at 28.5 C until dissociation,tissue:Lateral Ectoderm|genotype:WT TgSebox::eGFP embryos|treatment:1 cell stage injection of PAmCherry1 mRNA  Photoactivation  FACS,GSM7989705,GSM7989705: Zebrafish Lateral Ectoderm 7.5 hpf 1; Danio rerio; RNA Seq,GSM7989705 r1,GSM7989705,1,Cells were subjected to RNA extraction using RNeasy Plus Micro Kit QIAGEN according to the manufacturer instructions Complete cDNA synthesis and library preparation were performed using SMART Seq v3 protocol with Nextera UDI adapters. Libraries were then quantified by qPCR KAPA Biosysytems.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina NovaSeq 6000,,SRP479683,,,Lateral_Ectoderm_1.mate2.fastq.gz Lateral_Ectoderm_1.mate1.fastq.gz,fastq fastq,10106832046.0,43799441.0,GSM7989705 r1,0:80.75 1:150,A:3384386532;C:1698196797;G:2308026869;T:2706071735;N:10150113,80,150,,,3384386532,1698196797,2308026869,2706071735,10150113,SRX23006266,SRS19970079,SRA1774740,"Heisenberg group, Institute of Science and Technology Austria (ISTA)","Heisenberg group, Institute of Science and Technology Austria (ISTA)",2,0.64457,0.42542,0.04802,0.01964,0.91541,0.93878,0.83455,0.83564,68,150,B,B,mate2-mate1 similar by mapping diff,illumina,novaseq_era,unknown,cdna_unspecified,nextera,sc,single_cell_plate,smartseq,,Austria,2023-12-22,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
29206,SRR27329261,SRX23006265,SRS19970077,SRP479683,PRJNA1055904,BMP dependent patterning of ectoderm tissue material properties modulates lateral mesendoderm cell migration during early zebrafish gastrulation,GSE251904,Transcriptome Analysis,Cell migration is a fundamental process during embryonic development. Most studies in vivo have focussed on the migration of cells using the extracellular matrix ECM as their substrate for migration. In contrast  much less is known about how cells migrate on other cells  as found in early embryos when the ECM has not yet formed. Here  we show that lateral mesendoderm LME cells in the early zebrafish gastrula use the ectoderm as their substrate for migration. We show that the lateral ectoderm is permissive for the animal pole directed migration of LME cells  while the ectoderm at the animal pole halts it. These differences in the permissiveness depend on the lateral ectoderm being more cohesive than the animal ectoderm  a property controlled by BMP signalling within the ectoderm. Collectively  these findings identify ectoderm tissue cohesion as one critical factor regulating LME migration during zebrafish gastrulation. Overall design: To investigate the differences in gene expression between animal and lateral ectoderm causing the difference in material properties between the two ectodermal tissues.,,pubmed:40057955,,Zebrafish Animal Ectoderm 7.5 hpf 4,GSM7989704,,source name:Animal Ectoderm|tissue:Animal Ectoderm|genotype:WT TgSebox::eGFP embryos|treatment:1 cell stage injection of PAmCherry1 mRNA  Photoactivation  FACS|geo loc name:missing|collection date:missing,Zebrafish Animal Ectoderm 7.5 hpf 4,FastQC quality control Trimmomatic adapter & quality trimming; parameters: 2:30:10 LEADING:3 TRAILING:3 SLIDINGWINDOW:4:15 MINLEN:19 FastQC controlling trimming results salmon Alignment; used flags:   seqBias   qcBias R   DESeq2 differential gene expression analysis; DESEq2  apeglm  tximport  genefilter  GenomicRanges R   EnhancedVolcano & biomaRt visualization Assembly: Danio rerio GRCz11 Supplementary files format and content: tab delimited text file with gene name  gene length  effective length  tpm and number of reads for each sample,Animal Ectoderm,Embryo were injected with 125 pg of PAmCherry1 mRNA at xxx cell stage. At 6 hpf  embryos were mounted for up right imaging and a 444 x 220 x 58.5 µm volume was photoactivated using a 405 nm laser either at the animal pole or on the lateral side. The photoactivation process was composed of 10 cycles of 28 sec per embryo in a time window of approximately 90 min. post photoactivation  embryos were transferred in Ca2+ free Ringer solution and dissociated by gently pipetting using a P 1000 pipette to obtain a single cell suspension. 5000 PAmCherry1+ cells were sorted in 150 µl of RTL plus lysis buffer RNeasy Plus Micro Kit  QIAGEN with 1% of 2 mercaptoethanol.,Cells were subjected to RNA extraction using RNeasy Plus Micro Kit QIAGEN according to the manufacturer instructions Complete cDNA synthesis and library preparation were performed using SMART Seq v3 protocol with Nextera UDI adapters. Libraries were then quantified by qPCR KAPA Biosysytems.,Embryos were kept at 28.5 C until dissociation,tissue:Animal Ectoderm|genotype:WT TgSebox::eGFP embryos|treatment:1 cell stage injection of PAmCherry1 mRNA  Photoactivation  FACS,GSM7989704,GSM7989704: Zebrafish Animal Ectoderm 7.5 hpf 4; Danio rerio; RNA Seq,GSM7989704 r1,GSM7989704,1,Cells were subjected to RNA extraction using RNeasy Plus Micro Kit QIAGEN according to the manufacturer instructions Complete cDNA synthesis and library preparation were performed using SMART Seq v3 protocol with Nextera UDI adapters. Libraries were then quantified by qPCR KAPA Biosysytems.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina NovaSeq 6000,,SRP479683,,,Animal_Ectoderm_4.mate1.fastq.gz Animal_Ectoderm_4.mate2.fastq.gz,fastq fastq,5882974802.0,25612144.0,GSM7989704 r1,0:79.69 1:150.00,A:1784390197;C:1112846767;G:1392719296;T:1587227818;N:5790724,79,150,,,1784390197,1112846767,1392719296,1587227818,5790724,SRX23006265,SRS19970077,SRA1774740,"Heisenberg group, Institute of Science and Technology Austria (ISTA)","Heisenberg group, Institute of Science and Technology Austria (ISTA)",2,0.72169,0.51947,0.04749,0.0244,0.89061,0.91007,0.76259,0.75355,90,150,B,B,mate2-mate1 similar by mapping diff,illumina,novaseq_era,unknown,cdna_unspecified,nextera,sc,single_cell_plate,smartseq,,Austria,2023-12-22,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
29207,SRR27329262,SRX23006264,SRS19970076,SRP479683,PRJNA1055904,BMP dependent patterning of ectoderm tissue material properties modulates lateral mesendoderm cell migration during early zebrafish gastrulation,GSE251904,Transcriptome Analysis,Cell migration is a fundamental process during embryonic development. Most studies in vivo have focussed on the migration of cells using the extracellular matrix ECM as their substrate for migration. In contrast  much less is known about how cells migrate on other cells  as found in early embryos when the ECM has not yet formed. Here  we show that lateral mesendoderm LME cells in the early zebrafish gastrula use the ectoderm as their substrate for migration. We show that the lateral ectoderm is permissive for the animal pole directed migration of LME cells  while the ectoderm at the animal pole halts it. These differences in the permissiveness depend on the lateral ectoderm being more cohesive than the animal ectoderm  a property controlled by BMP signalling within the ectoderm. Collectively  these findings identify ectoderm tissue cohesion as one critical factor regulating LME migration during zebrafish gastrulation. Overall design: To investigate the differences in gene expression between animal and lateral ectoderm causing the difference in material properties between the two ectodermal tissues.,,pubmed:40057955,,Zebrafish Animal Ectoderm 7.5 hpf 3,GSM7989703,,source name:Animal Ectoderm|tissue:Animal Ectoderm|genotype:WT TgSebox::eGFP embryos|treatment:1 cell stage injection of PAmCherry1 mRNA  Photoactivation  FACS|geo loc name:missing|collection date:missing,Zebrafish Animal Ectoderm 7.5 hpf 3,FastQC quality control Trimmomatic adapter & quality trimming; parameters: 2:30:10 LEADING:3 TRAILING:3 SLIDINGWINDOW:4:15 MINLEN:19 FastQC controlling trimming results salmon Alignment; used flags:   seqBias   qcBias R   DESeq2 differential gene expression analysis; DESEq2  apeglm  tximport  genefilter  GenomicRanges R   EnhancedVolcano & biomaRt visualization Assembly: Danio rerio GRCz11 Supplementary files format and content: tab delimited text file with gene name  gene length  effective length  tpm and number of reads for each sample,Animal Ectoderm,Embryo were injected with 125 pg of PAmCherry1 mRNA at xxx cell stage. At 6 hpf  embryos were mounted for up right imaging and a 444 x 220 x 58.5 µm volume was photoactivated using a 405 nm laser either at the animal pole or on the lateral side. The photoactivation process was composed of 10 cycles of 28 sec per embryo in a time window of approximately 90 min. post photoactivation  embryos were transferred in Ca2+ free Ringer solution and dissociated by gently pipetting using a P 1000 pipette to obtain a single cell suspension. 5000 PAmCherry1+ cells were sorted in 150 µl of RTL plus lysis buffer RNeasy Plus Micro Kit  QIAGEN with 1% of 2 mercaptoethanol.,Cells were subjected to RNA extraction using RNeasy Plus Micro Kit QIAGEN according to the manufacturer instructions Complete cDNA synthesis and library preparation were performed using SMART Seq v3 protocol with Nextera UDI adapters. Libraries were then quantified by qPCR KAPA Biosysytems.,Embryos were kept at 28.5 C until dissociation,tissue:Animal Ectoderm|genotype:WT TgSebox::eGFP embryos|treatment:1 cell stage injection of PAmCherry1 mRNA  Photoactivation  FACS,GSM7989703,GSM7989703: Zebrafish Animal Ectoderm 7.5 hpf 3; Danio rerio; RNA Seq,GSM7989703 r1,GSM7989703,1,Cells were subjected to RNA extraction using RNeasy Plus Micro Kit QIAGEN according to the manufacturer instructions Complete cDNA synthesis and library preparation were performed using SMART Seq v3 protocol with Nextera UDI adapters. Libraries were then quantified by qPCR KAPA Biosysytems.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina NovaSeq 6000,,SRP479683,,,Animal_Ectoderm_3.mate1.fastq.gz Animal_Ectoderm_3.mate2.fastq.gz,fastq fastq,6802628678.0,29741168.0,GSM7989703 r1,0:78.73 1:150,A:1935418861;C:1362666541;G:1651315180;T:1846520380;N:6707716,78,150,,,1935418861,1362666541,1651315180,1846520380,6707716,SRX23006264,SRS19970076,SRA1774740,"Heisenberg group, Institute of Science and Technology Austria (ISTA)","Heisenberg group, Institute of Science and Technology Austria (ISTA)",2,0.74265,0.48713,0.04983,0.02636,0.87771,0.90335,0.72388,0.71557,68,150,B,B,mate2-mate1 similar by mapping diff,illumina,novaseq_era,unknown,cdna_unspecified,nextera,sc,single_cell_plate,smartseq,,Austria,2023-12-22,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
29208,SRR27329263,SRX23006263,SRS19970075,SRP479683,PRJNA1055904,BMP dependent patterning of ectoderm tissue material properties modulates lateral mesendoderm cell migration during early zebrafish gastrulation,GSE251904,Transcriptome Analysis,Cell migration is a fundamental process during embryonic development. Most studies in vivo have focussed on the migration of cells using the extracellular matrix ECM as their substrate for migration. In contrast  much less is known about how cells migrate on other cells  as found in early embryos when the ECM has not yet formed. Here  we show that lateral mesendoderm LME cells in the early zebrafish gastrula use the ectoderm as their substrate for migration. We show that the lateral ectoderm is permissive for the animal pole directed migration of LME cells  while the ectoderm at the animal pole halts it. These differences in the permissiveness depend on the lateral ectoderm being more cohesive than the animal ectoderm  a property controlled by BMP signalling within the ectoderm. Collectively  these findings identify ectoderm tissue cohesion as one critical factor regulating LME migration during zebrafish gastrulation. Overall design: To investigate the differences in gene expression between animal and lateral ectoderm causing the difference in material properties between the two ectodermal tissues.,,pubmed:40057955,,Zebrafish Animal Ectoderm 7.5 hpf 2,GSM7989702,,source name:Animal Ectoderm|tissue:Animal Ectoderm|genotype:WT TgSebox::eGFP embryos|treatment:1 cell stage injection of PAmCherry1 mRNA  Photoactivation  FACS|geo loc name:missing|collection date:missing,Zebrafish Animal Ectoderm 7.5 hpf 2,FastQC quality control Trimmomatic adapter & quality trimming; parameters: 2:30:10 LEADING:3 TRAILING:3 SLIDINGWINDOW:4:15 MINLEN:19 FastQC controlling trimming results salmon Alignment; used flags:   seqBias   qcBias R   DESeq2 differential gene expression analysis; DESEq2  apeglm  tximport  genefilter  GenomicRanges R   EnhancedVolcano & biomaRt visualization Assembly: Danio rerio GRCz11 Supplementary files format and content: tab delimited text file with gene name  gene length  effective length  tpm and number of reads for each sample,Animal Ectoderm,Embryo were injected with 125 pg of PAmCherry1 mRNA at xxx cell stage. At 6 hpf  embryos were mounted for up right imaging and a 444 x 220 x 58.5 µm volume was photoactivated using a 405 nm laser either at the animal pole or on the lateral side. The photoactivation process was composed of 10 cycles of 28 sec per embryo in a time window of approximately 90 min. post photoactivation  embryos were transferred in Ca2+ free Ringer solution and dissociated by gently pipetting using a P 1000 pipette to obtain a single cell suspension. 5000 PAmCherry1+ cells were sorted in 150 µl of RTL plus lysis buffer RNeasy Plus Micro Kit  QIAGEN with 1% of 2 mercaptoethanol.,Cells were subjected to RNA extraction using RNeasy Plus Micro Kit QIAGEN according to the manufacturer instructions Complete cDNA synthesis and library preparation were performed using SMART Seq v3 protocol with Nextera UDI adapters. Libraries were then quantified by qPCR KAPA Biosysytems.,Embryos were kept at 28.5 C until dissociation,tissue:Animal Ectoderm|genotype:WT TgSebox::eGFP embryos|treatment:1 cell stage injection of PAmCherry1 mRNA  Photoactivation  FACS,GSM7989702,GSM7989702: Zebrafish Animal Ectoderm 7.5 hpf 2; Danio rerio; RNA Seq,GSM7989702 r1,GSM7989702,1,Cells were subjected to RNA extraction using RNeasy Plus Micro Kit QIAGEN according to the manufacturer instructions Complete cDNA synthesis and library preparation were performed using SMART Seq v3 protocol with Nextera UDI adapters. Libraries were then quantified by qPCR KAPA Biosysytems.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina NovaSeq 6000,,SRP479683,,,Animal_Ectoderm_2.mate1.fastq.gz Animal_Ectoderm_2.mate2.fastq.gz,fastq fastq,6083410864.0,26183986.0,GSM7989702 r1,0:82.33 1:150,A:1976903262;C:1110974120;G:1376486325;T:1613990197;N:5056960,82,150,,,1976903262,1110974120,1376486325,1613990197,5056960,SRX23006263,SRS19970075,SRA1774740,"Heisenberg group, Institute of Science and Technology Austria (ISTA)","Heisenberg group, Institute of Science and Technology Austria (ISTA)",2,0.73078,0.55778,0.04052,0.02684,0.89509,0.91106,0.77391,0.7649,68,150,B,B,biological fallback assumption,illumina,novaseq_era,unknown,cdna_unspecified,nextera,sc,single_cell_plate,smartseq,,Austria,2023-12-22,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
29209,SRR27329264,SRX23006262,SRS19970074,SRP479683,PRJNA1055904,BMP dependent patterning of ectoderm tissue material properties modulates lateral mesendoderm cell migration during early zebrafish gastrulation,GSE251904,Transcriptome Analysis,Cell migration is a fundamental process during embryonic development. Most studies in vivo have focussed on the migration of cells using the extracellular matrix ECM as their substrate for migration. In contrast  much less is known about how cells migrate on other cells  as found in early embryos when the ECM has not yet formed. Here  we show that lateral mesendoderm LME cells in the early zebrafish gastrula use the ectoderm as their substrate for migration. We show that the lateral ectoderm is permissive for the animal pole directed migration of LME cells  while the ectoderm at the animal pole halts it. These differences in the permissiveness depend on the lateral ectoderm being more cohesive than the animal ectoderm  a property controlled by BMP signalling within the ectoderm. Collectively  these findings identify ectoderm tissue cohesion as one critical factor regulating LME migration during zebrafish gastrulation. Overall design: To investigate the differences in gene expression between animal and lateral ectoderm causing the difference in material properties between the two ectodermal tissues.,,pubmed:40057955,,Zebrafish Animal Ectoderm 7.5 hpf 1,GSM7989701,,source name:Animal Ectoderm|tissue:Animal Ectoderm|genotype:WT TgSebox::eGFP embryos|treatment:1 cell stage injection of PAmCherry1 mRNA  Photoactivation  FACS|geo loc name:missing|collection date:missing,Zebrafish Animal Ectoderm 7.5 hpf 1,FastQC quality control Trimmomatic adapter & quality trimming; parameters: 2:30:10 LEADING:3 TRAILING:3 SLIDINGWINDOW:4:15 MINLEN:19 FastQC controlling trimming results salmon Alignment; used flags:   seqBias   qcBias R   DESeq2 differential gene expression analysis; DESEq2  apeglm  tximport  genefilter  GenomicRanges R   EnhancedVolcano & biomaRt visualization Assembly: Danio rerio GRCz11 Supplementary files format and content: tab delimited text file with gene name  gene length  effective length  tpm and number of reads for each sample,Animal Ectoderm,Embryo were injected with 125 pg of PAmCherry1 mRNA at xxx cell stage. At 6 hpf  embryos were mounted for up right imaging and a 444 x 220 x 58.5 µm volume was photoactivated using a 405 nm laser either at the animal pole or on the lateral side. The photoactivation process was composed of 10 cycles of 28 sec per embryo in a time window of approximately 90 min. post photoactivation  embryos were transferred in Ca2+ free Ringer solution and dissociated by gently pipetting using a P 1000 pipette to obtain a single cell suspension. 5000 PAmCherry1+ cells were sorted in 150 µl of RTL plus lysis buffer RNeasy Plus Micro Kit  QIAGEN with 1% of 2 mercaptoethanol.,Cells were subjected to RNA extraction using RNeasy Plus Micro Kit QIAGEN according to the manufacturer instructions Complete cDNA synthesis and library preparation were performed using SMART Seq v3 protocol with Nextera UDI adapters. Libraries were then quantified by qPCR KAPA Biosysytems.,Embryos were kept at 28.5 C until dissociation,tissue:Animal Ectoderm|genotype:WT TgSebox::eGFP embryos|treatment:1 cell stage injection of PAmCherry1 mRNA  Photoactivation  FACS,GSM7989701,GSM7989701: Zebrafish Animal Ectoderm 7.5 hpf 1; Danio rerio; RNA Seq,GSM7989701 r1,GSM7989701,1,Cells were subjected to RNA extraction using RNeasy Plus Micro Kit QIAGEN according to the manufacturer instructions Complete cDNA synthesis and library preparation were performed using SMART Seq v3 protocol with Nextera UDI adapters. Libraries were then quantified by qPCR KAPA Biosysytems.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina NovaSeq 6000,,SRP479683,,,Animal_Ectoderm_1.mate1.fastq.gz Animal_Ectoderm_1.mate2.fastq.gz,fastq fastq,7195859200.0,31306406.0,GSM7989701 r1,0:79.85 1:150,A:2025250331;C:1450062111;G:1771726396;T:1941745747;N:7074615,79,150,,,2025250331,1450062111,1771726396,1941745747,7074615,SRX23006262,SRS19970074,SRA1774740,"Heisenberg group, Institute of Science and Technology Austria (ISTA)","Heisenberg group, Institute of Science and Technology Austria (ISTA)",2,0.74201,0.52197,0.04312,0.02595,0.87941,0.9013,0.74298,0.71509,68,150,B,B,mate2-mate1 similar by mapping diff,illumina,novaseq_era,unknown,cdna_unspecified,nextera,sc,single_cell_plate,smartseq,,Austria,2023-12-22,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
44076,SRR6250374,SRX3357301,SRS2656385,SRP123526,PRJNA416939,Single cell RNAseq SMART seq2 of wild type TLAB and MZoep tz57 zebrafish embryos at 50% epiboly stage,GSE106466,Transcriptome Analysis,SMART seq2 was performed on single cells isolated from visually staged zebrafish embryos. Overall design: Samples were all sequenced in one batch. Some were generated with a five prime' UMI tagged method  and others are full length SMART seq2.,parent bioproject:PRJNA417291,pubmed:29700225,,wt p4 S9,GSM2838460,,tissue:Embryonic cell|developmental stage:50% epiboly|library preparation batch:p4|genotype:Wild type TLAB,wt p4 S9,Bases 1 8 the UMI and template switch were trimmed from the remaining reads  trimmed stretches of polyA or polyT longer than 6bp and removed any reads that were shorter than 15 bp post trimming. We identified and depleted mtRNA and rRNA reads from the remaining reads by using Bowtie2 to align all reads against a transcriptome of mtRNA and rRNA  using the parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed   no unal. Remaining reads were then aligned to a modified version of the Ensembl Zv10 release 81 zebrafish transcriptome using Bowtie2 with parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed. The transcriptome start site of each gene was padded by 100bp to increase alignment rates  since reads were generated from the 5’ end of the transcript  and not all transcripts had correct annotations for their 5’ end. Additionally  pseudogenes were removed from the transcriptome as they sometimes interfered with real genes  given our 5’ coordinate selection described below. We chose the best alignment per read as the read with the most 5’ alignment coordinate to a transcript among those that aligned to the correct strand and had an alignment score of no more than 4 below the best alignment score overall. We then iteratively collapsed pairs of UMIs  if the sum of the minimum Phred quality scores across non matching bases was at most 30 to account for mis matched bases that were likely due to sequencing errors. A collapsed pair was represented for future comparisons by the UMI with a larger prior number of aligned reads unless the other UMI had at least 75% as many reads aligned and had equal or better UMI quality scores for those alignments evaluated as the softmin over aligned reads of the softmax over quality scores for the UMI bases in an aligned read. post collapsing  the number of transcripts observed per gene was calculated. This value was then adjusted based on the probability of a collision the drawing at random of the same UMI multiple times  which was calculated from the number of UMIs not detected for a given gene. The full length fragmented reads without xxx’ enrichment or UMI were aligned using RSEM against a reference transcriptome  which is depleted of mtRNA  rRNA  and pseudogenes as described above. The command ‘rsem calculate expression’ was used with parameters  p 4   paired end   seed length 24 for performing the alignment. The resulting gene levels in FPKM was used to build the expression matrix. Genome build: GRCz10 dr81 Supplementary files format and content: Tab delimited text files include gene levels UMI counts or FKPM for each sample.,Embryonic cell,,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for 5’ enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,Fish of each genotype were incrossed and embryos were collected. Embryos were dechorionated and allowed to grow at 28°C in standard fish water containing methylene blue till 50% epiboly stage.,developmental stage:50% epiboly|library preparation batch:p4|genotype:Wild type TLAB,GSM2838460,GSM2838460: wt p4 S9; Danio rerio; RNA Seq,GSM2838460,,1,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for five prime enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina HiSeq 2500,,SRP123526,,,wt_p4_S9.R1.fastq.gz wt_p4_S9.R2.fastq.gz,fastq fastq,61039160.0,1089985.0,GSM2838460 r1,0:24 1:32,A:16411881;C:13934629;G:14159762;T:16484450;N:48438,24,32,,,16411881,13934629,14159762,16484450,48438,SRX3357301,SRS2656385,SRA627905,GEO,"Schier Lab, Molecular and Cellular Biology, Harvard University",2,0.77476,0.82366,0.12876,0.13588,0.85766,0.85774,0.72622,0.72762,24,32,B,B,biological fallback assumption,illumina,hiseq_era,full_length,poly_a,unknown,sc,single_cell_plate,smartseq,,United States,2017-11-02,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
44077,SRR6250373,SRX3357300,SRS2656381,SRP123526,PRJNA416939,Single cell RNAseq SMART seq2 of wild type TLAB and MZoep tz57 zebrafish embryos at 50% epiboly stage,GSE106466,Transcriptome Analysis,SMART seq2 was performed on single cells isolated from visually staged zebrafish embryos. Overall design: Samples were all sequenced in one batch. Some were generated with a five prime' UMI tagged method  and others are full length SMART seq2.,parent bioproject:PRJNA417291,pubmed:29700225,,wt p4 S8,GSM2838459,,tissue:Embryonic cell|developmental stage:50% epiboly|library preparation batch:p4|genotype:Wild type TLAB,wt p4 S8,Bases 1 8 the UMI and template switch were trimmed from the remaining reads  trimmed stretches of polyA or polyT longer than 6bp and removed any reads that were shorter than 15 bp post trimming. We identified and depleted mtRNA and rRNA reads from the remaining reads by using Bowtie2 to align all reads against a transcriptome of mtRNA and rRNA  using the parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed   no unal. Remaining reads were then aligned to a modified version of the Ensembl Zv10 release 81 zebrafish transcriptome using Bowtie2 with parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed. The transcriptome start site of each gene was padded by 100bp to increase alignment rates  since reads were generated from the 5’ end of the transcript  and not all transcripts had correct annotations for their 5’ end. Additionally  pseudogenes were removed from the transcriptome as they sometimes interfered with real genes  given our 5’ coordinate selection described below. We chose the best alignment per read as the read with the most 5’ alignment coordinate to a transcript among those that aligned to the correct strand and had an alignment score of no more than 4 below the best alignment score overall. We then iteratively collapsed pairs of UMIs  if the sum of the minimum Phred quality scores across non matching bases was at most 30 to account for mis matched bases that were likely due to sequencing errors. A collapsed pair was represented for future comparisons by the UMI with a larger prior number of aligned reads unless the other UMI had at least 75% as many reads aligned and had equal or better UMI quality scores for those alignments evaluated as the softmin over aligned reads of the softmax over quality scores for the UMI bases in an aligned read. post collapsing  the number of transcripts observed per gene was calculated. This value was then adjusted based on the probability of a collision the drawing at random of the same UMI multiple times  which was calculated from the number of UMIs not detected for a given gene. The full length fragmented reads without xxx’ enrichment or UMI were aligned using RSEM against a reference transcriptome  which is depleted of mtRNA  rRNA  and pseudogenes as described above. The command ‘rsem calculate expression’ was used with parameters  p 4   paired end   seed length 24 for performing the alignment. The resulting gene levels in FPKM was used to build the expression matrix. Genome build: GRCz10 dr81 Supplementary files format and content: Tab delimited text files include gene levels UMI counts or FKPM for each sample.,Embryonic cell,,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for 5’ enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,Fish of each genotype were incrossed and embryos were collected. Embryos were dechorionated and allowed to grow at 28°C in standard fish water containing methylene blue till 50% epiboly stage.,developmental stage:50% epiboly|library preparation batch:p4|genotype:Wild type TLAB,GSM2838459,GSM2838459: wt p4 S8; Danio rerio; RNA Seq,GSM2838459,,1,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for five prime enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina HiSeq 2500,,SRP123526,,,wt_p4_S8.R1.fastq.gz wt_p4_S8.R2.fastq.gz,fastq fastq,3221400.0,57525.0,GSM2838459 r1,0:24 1:32,A:857299;C:737999;G:752784;T:870887;N:2431,24,32,,,857299,737999,752784,870887,2431,SRX3357300,SRS2656381,SRA627905,GEO,"Schier Lab, Molecular and Cellular Biology, Harvard University",2,0.76405,0.81832,0.15327,0.16586,0.90983,0.90826,0.73965,0.74051,24,32,B,B,biological fallback assumption,illumina,hiseq_era,full_length,poly_a,unknown,sc,single_cell_plate,smartseq,,United States,2017-11-02,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
44078,SRR6250372,SRX3357299,SRS2656383,SRP123526,PRJNA416939,Single cell RNAseq SMART seq2 of wild type TLAB and MZoep tz57 zebrafish embryos at 50% epiboly stage,GSE106466,Transcriptome Analysis,SMART seq2 was performed on single cells isolated from visually staged zebrafish embryos. Overall design: Samples were all sequenced in one batch. Some were generated with a five prime' UMI tagged method  and others are full length SMART seq2.,parent bioproject:PRJNA417291,pubmed:29700225,,wt p4 S7,GSM2838458,,tissue:Embryonic cell|developmental stage:50% epiboly|library preparation batch:p4|genotype:Wild type TLAB,wt p4 S7,Bases 1 8 the UMI and template switch were trimmed from the remaining reads  trimmed stretches of polyA or polyT longer than 6bp and removed any reads that were shorter than 15 bp post trimming. We identified and depleted mtRNA and rRNA reads from the remaining reads by using Bowtie2 to align all reads against a transcriptome of mtRNA and rRNA  using the parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed   no unal. Remaining reads were then aligned to a modified version of the Ensembl Zv10 release 81 zebrafish transcriptome using Bowtie2 with parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed. The transcriptome start site of each gene was padded by 100bp to increase alignment rates  since reads were generated from the 5’ end of the transcript  and not all transcripts had correct annotations for their 5’ end. Additionally  pseudogenes were removed from the transcriptome as they sometimes interfered with real genes  given our 5’ coordinate selection described below. We chose the best alignment per read as the read with the most 5’ alignment coordinate to a transcript among those that aligned to the correct strand and had an alignment score of no more than 4 below the best alignment score overall. We then iteratively collapsed pairs of UMIs  if the sum of the minimum Phred quality scores across non matching bases was at most 30 to account for mis matched bases that were likely due to sequencing errors. A collapsed pair was represented for future comparisons by the UMI with a larger prior number of aligned reads unless the other UMI had at least 75% as many reads aligned and had equal or better UMI quality scores for those alignments evaluated as the softmin over aligned reads of the softmax over quality scores for the UMI bases in an aligned read. post collapsing  the number of transcripts observed per gene was calculated. This value was then adjusted based on the probability of a collision the drawing at random of the same UMI multiple times  which was calculated from the number of UMIs not detected for a given gene. The full length fragmented reads without xxx’ enrichment or UMI were aligned using RSEM against a reference transcriptome  which is depleted of mtRNA  rRNA  and pseudogenes as described above. The command ‘rsem calculate expression’ was used with parameters  p 4   paired end   seed length 24 for performing the alignment. The resulting gene levels in FPKM was used to build the expression matrix. Genome build: GRCz10 dr81 Supplementary files format and content: Tab delimited text files include gene levels UMI counts or FKPM for each sample.,Embryonic cell,,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for 5’ enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,Fish of each genotype were incrossed and embryos were collected. Embryos were dechorionated and allowed to grow at 28°C in standard fish water containing methylene blue till 50% epiboly stage.,developmental stage:50% epiboly|library preparation batch:p4|genotype:Wild type TLAB,GSM2838458,GSM2838458: wt p4 S7; Danio rerio; RNA Seq,GSM2838458,,1,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for five prime enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina HiSeq 2500,,SRP123526,,,wt_p4_S7.R1.fastq.gz wt_p4_S7.R2.fastq.gz,fastq fastq,61365472.0,1095812.0,GSM2838458 r1,0:24 1:32,A:16777185;C:13827520;G:13996928;T:16714262;N:49577,24,32,,,16777185,13827520,13996928,16714262,49577,SRX3357299,SRS2656383,SRA627905,GEO,"Schier Lab, Molecular and Cellular Biology, Harvard University",2,0.78049,0.82981,0.09421,0.09809,0.8579,0.85882,0.74187,0.74992,24,32,B,B,biological fallback assumption,illumina,hiseq_era,full_length,poly_a,unknown,sc,single_cell_plate,smartseq,,United States,2017-11-02,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
44079,SRR6250371,SRX3357298,SRS2656380,SRP123526,PRJNA416939,Single cell RNAseq SMART seq2 of wild type TLAB and MZoep tz57 zebrafish embryos at 50% epiboly stage,GSE106466,Transcriptome Analysis,SMART seq2 was performed on single cells isolated from visually staged zebrafish embryos. Overall design: Samples were all sequenced in one batch. Some were generated with a five prime' UMI tagged method  and others are full length SMART seq2.,parent bioproject:PRJNA417291,pubmed:29700225,,wt p4 S6,GSM2838457,,tissue:Embryonic cell|developmental stage:50% epiboly|library preparation batch:p4|genotype:Wild type TLAB,wt p4 S6,Bases 1 8 the UMI and template switch were trimmed from the remaining reads  trimmed stretches of polyA or polyT longer than 6bp and removed any reads that were shorter than 15 bp post trimming. We identified and depleted mtRNA and rRNA reads from the remaining reads by using Bowtie2 to align all reads against a transcriptome of mtRNA and rRNA  using the parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed   no unal. Remaining reads were then aligned to a modified version of the Ensembl Zv10 release 81 zebrafish transcriptome using Bowtie2 with parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed. The transcriptome start site of each gene was padded by 100bp to increase alignment rates  since reads were generated from the 5’ end of the transcript  and not all transcripts had correct annotations for their 5’ end. Additionally  pseudogenes were removed from the transcriptome as they sometimes interfered with real genes  given our 5’ coordinate selection described below. We chose the best alignment per read as the read with the most 5’ alignment coordinate to a transcript among those that aligned to the correct strand and had an alignment score of no more than 4 below the best alignment score overall. We then iteratively collapsed pairs of UMIs  if the sum of the minimum Phred quality scores across non matching bases was at most 30 to account for mis matched bases that were likely due to sequencing errors. A collapsed pair was represented for future comparisons by the UMI with a larger prior number of aligned reads unless the other UMI had at least 75% as many reads aligned and had equal or better UMI quality scores for those alignments evaluated as the softmin over aligned reads of the softmax over quality scores for the UMI bases in an aligned read. post collapsing  the number of transcripts observed per gene was calculated. This value was then adjusted based on the probability of a collision the drawing at random of the same UMI multiple times  which was calculated from the number of UMIs not detected for a given gene. The full length fragmented reads without xxx’ enrichment or UMI were aligned using RSEM against a reference transcriptome  which is depleted of mtRNA  rRNA  and pseudogenes as described above. The command ‘rsem calculate expression’ was used with parameters  p 4   paired end   seed length 24 for performing the alignment. The resulting gene levels in FPKM was used to build the expression matrix. Genome build: GRCz10 dr81 Supplementary files format and content: Tab delimited text files include gene levels UMI counts or FKPM for each sample.,Embryonic cell,,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for 5’ enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,Fish of each genotype were incrossed and embryos were collected. Embryos were dechorionated and allowed to grow at 28°C in standard fish water containing methylene blue till 50% epiboly stage.,developmental stage:50% epiboly|library preparation batch:p4|genotype:Wild type TLAB,GSM2838457,GSM2838457: wt p4 S6; Danio rerio; RNA Seq,GSM2838457,,1,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for five prime enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina HiSeq 2500,,SRP123526,,,wt_p4_S6.R1.fastq.gz wt_p4_S6.R2.fastq.gz,fastq fastq,70813008.0,1264518.0,GSM2838457 r1,0:24 1:32,A:18473405;C:16677041;G:16946924;T:18657517;N:58121,24,32,,,18473405,16677041,16946924,18657517,58121,SRX3357298,SRS2656380,SRA627905,GEO,"Schier Lab, Molecular and Cellular Biology, Harvard University",2,0.78649,0.83922,0.17937,0.19426,0.84656,0.84713,0.69099,0.68701,24,32,B,B,biological fallback assumption,illumina,hiseq_era,full_length,poly_a,unknown,sc,single_cell_plate,smartseq,,United States,2017-11-02,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
44080,SRR6250370,SRX3357296,SRS2656379,SRP123526,PRJNA416939,Single cell RNAseq SMART seq2 of wild type TLAB and MZoep tz57 zebrafish embryos at 50% epiboly stage,GSE106466,Transcriptome Analysis,SMART seq2 was performed on single cells isolated from visually staged zebrafish embryos. Overall design: Samples were all sequenced in one batch. Some were generated with a five prime' UMI tagged method  and others are full length SMART seq2.,parent bioproject:PRJNA417291,pubmed:29700225,,wt p4 S5,GSM2838456,,tissue:Embryonic cell|developmental stage:50% epiboly|library preparation batch:p4|genotype:Wild type TLAB,wt p4 S5,Bases 1 8 the UMI and template switch were trimmed from the remaining reads  trimmed stretches of polyA or polyT longer than 6bp and removed any reads that were shorter than 15 bp post trimming. We identified and depleted mtRNA and rRNA reads from the remaining reads by using Bowtie2 to align all reads against a transcriptome of mtRNA and rRNA  using the parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed   no unal. Remaining reads were then aligned to a modified version of the Ensembl Zv10 release 81 zebrafish transcriptome using Bowtie2 with parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed. The transcriptome start site of each gene was padded by 100bp to increase alignment rates  since reads were generated from the 5’ end of the transcript  and not all transcripts had correct annotations for their 5’ end. Additionally  pseudogenes were removed from the transcriptome as they sometimes interfered with real genes  given our 5’ coordinate selection described below. We chose the best alignment per read as the read with the most 5’ alignment coordinate to a transcript among those that aligned to the correct strand and had an alignment score of no more than 4 below the best alignment score overall. We then iteratively collapsed pairs of UMIs  if the sum of the minimum Phred quality scores across non matching bases was at most 30 to account for mis matched bases that were likely due to sequencing errors. A collapsed pair was represented for future comparisons by the UMI with a larger prior number of aligned reads unless the other UMI had at least 75% as many reads aligned and had equal or better UMI quality scores for those alignments evaluated as the softmin over aligned reads of the softmax over quality scores for the UMI bases in an aligned read. post collapsing  the number of transcripts observed per gene was calculated. This value was then adjusted based on the probability of a collision the drawing at random of the same UMI multiple times  which was calculated from the number of UMIs not detected for a given gene. The full length fragmented reads without xxx’ enrichment or UMI were aligned using RSEM against a reference transcriptome  which is depleted of mtRNA  rRNA  and pseudogenes as described above. The command ‘rsem calculate expression’ was used with parameters  p 4   paired end   seed length 24 for performing the alignment. The resulting gene levels in FPKM was used to build the expression matrix. Genome build: GRCz10 dr81 Supplementary files format and content: Tab delimited text files include gene levels UMI counts or FKPM for each sample.,Embryonic cell,,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for 5’ enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,Fish of each genotype were incrossed and embryos were collected. Embryos were dechorionated and allowed to grow at 28°C in standard fish water containing methylene blue till 50% epiboly stage.,developmental stage:50% epiboly|library preparation batch:p4|genotype:Wild type TLAB,GSM2838456,GSM2838456: wt p4 S5; Danio rerio; RNA Seq,GSM2838456,,1,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for five prime enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina HiSeq 2500,,SRP123526,,,wt_p4_S5.R1.fastq.gz wt_p4_S5.R2.fastq.gz,fastq fastq,73596656.0,1314226.0,GSM2838456 r1,0:24 1:32,A:18627101;C:17852449;G:18124666;T:18933645;N:58795,24,32,,,18627101,17852449,18124666,18933645,58795,SRX3357296,SRS2656379,SRA627905,GEO,"Schier Lab, Molecular and Cellular Biology, Harvard University",2,0.78607,0.84024,0.21328,0.23201,0.85543,0.85405,0.69857,0.70151,24,32,B,B,biological fallback assumption,illumina,hiseq_era,full_length,poly_a,unknown,sc,single_cell_plate,smartseq,,United States,2017-11-02,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
44081,SRR6250369,SRX3357295,SRS2656377,SRP123526,PRJNA416939,Single cell RNAseq SMART seq2 of wild type TLAB and MZoep tz57 zebrafish embryos at 50% epiboly stage,GSE106466,Transcriptome Analysis,SMART seq2 was performed on single cells isolated from visually staged zebrafish embryos. Overall design: Samples were all sequenced in one batch. Some were generated with a five prime' UMI tagged method  and others are full length SMART seq2.,parent bioproject:PRJNA417291,pubmed:29700225,,wt p4 S4,GSM2838455,,tissue:Embryonic cell|developmental stage:50% epiboly|library preparation batch:p4|genotype:Wild type TLAB,wt p4 S4,Bases 1 8 the UMI and template switch were trimmed from the remaining reads  trimmed stretches of polyA or polyT longer than 6bp and removed any reads that were shorter than 15 bp post trimming. We identified and depleted mtRNA and rRNA reads from the remaining reads by using Bowtie2 to align all reads against a transcriptome of mtRNA and rRNA  using the parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed   no unal. Remaining reads were then aligned to a modified version of the Ensembl Zv10 release 81 zebrafish transcriptome using Bowtie2 with parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed. The transcriptome start site of each gene was padded by 100bp to increase alignment rates  since reads were generated from the 5’ end of the transcript  and not all transcripts had correct annotations for their 5’ end. Additionally  pseudogenes were removed from the transcriptome as they sometimes interfered with real genes  given our 5’ coordinate selection described below. We chose the best alignment per read as the read with the most 5’ alignment coordinate to a transcript among those that aligned to the correct strand and had an alignment score of no more than 4 below the best alignment score overall. We then iteratively collapsed pairs of UMIs  if the sum of the minimum Phred quality scores across non matching bases was at most 30 to account for mis matched bases that were likely due to sequencing errors. A collapsed pair was represented for future comparisons by the UMI with a larger prior number of aligned reads unless the other UMI had at least 75% as many reads aligned and had equal or better UMI quality scores for those alignments evaluated as the softmin over aligned reads of the softmax over quality scores for the UMI bases in an aligned read. post collapsing  the number of transcripts observed per gene was calculated. This value was then adjusted based on the probability of a collision the drawing at random of the same UMI multiple times  which was calculated from the number of UMIs not detected for a given gene. The full length fragmented reads without xxx’ enrichment or UMI were aligned using RSEM against a reference transcriptome  which is depleted of mtRNA  rRNA  and pseudogenes as described above. The command ‘rsem calculate expression’ was used with parameters  p 4   paired end   seed length 24 for performing the alignment. The resulting gene levels in FPKM was used to build the expression matrix. Genome build: GRCz10 dr81 Supplementary files format and content: Tab delimited text files include gene levels UMI counts or FKPM for each sample.,Embryonic cell,,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for 5’ enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,Fish of each genotype were incrossed and embryos were collected. Embryos were dechorionated and allowed to grow at 28°C in standard fish water containing methylene blue till 50% epiboly stage.,developmental stage:50% epiboly|library preparation batch:p4|genotype:Wild type TLAB,GSM2838455,GSM2838455: wt p4 S4; Danio rerio; RNA Seq,GSM2838455,,1,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for five prime enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina HiSeq 2500,,SRP123526,,,wt_p4_S4.R2.fastq.gz wt_p4_S4.R1.fastq.gz,fastq fastq,3035032.0,54197.0,GSM2838455 r1,0:24 1:32,A:827352;C:678981;G:690970;T:835185;N:2544,24,32,,,827352,678981,690970,835185,2544,SRX3357295,SRS2656377,SRA627905,GEO,"Schier Lab, Molecular and Cellular Biology, Harvard University",2,0.77706,0.82883,0.10498,0.11133,0.90701,0.90603,0.75427,0.76069,24,32,B,B,biological fallback assumption,illumina,hiseq_era,full_length,poly_a,unknown,sc,single_cell_plate,smartseq,,United States,2017-11-02,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
44082,SRR6250368,SRX3357294,SRS2656376,SRP123526,PRJNA416939,Single cell RNAseq SMART seq2 of wild type TLAB and MZoep tz57 zebrafish embryos at 50% epiboly stage,GSE106466,Transcriptome Analysis,SMART seq2 was performed on single cells isolated from visually staged zebrafish embryos. Overall design: Samples were all sequenced in one batch. Some were generated with a five prime' UMI tagged method  and others are full length SMART seq2.,parent bioproject:PRJNA417291,pubmed:29700225,,wt p4 S3,GSM2838454,,tissue:Embryonic cell|developmental stage:50% epiboly|library preparation batch:p4|genotype:Wild type TLAB,wt p4 S3,Bases 1 8 the UMI and template switch were trimmed from the remaining reads  trimmed stretches of polyA or polyT longer than 6bp and removed any reads that were shorter than 15 bp post trimming. We identified and depleted mtRNA and rRNA reads from the remaining reads by using Bowtie2 to align all reads against a transcriptome of mtRNA and rRNA  using the parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed   no unal. Remaining reads were then aligned to a modified version of the Ensembl Zv10 release 81 zebrafish transcriptome using Bowtie2 with parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed. The transcriptome start site of each gene was padded by 100bp to increase alignment rates  since reads were generated from the 5’ end of the transcript  and not all transcripts had correct annotations for their 5’ end. Additionally  pseudogenes were removed from the transcriptome as they sometimes interfered with real genes  given our 5’ coordinate selection described below. We chose the best alignment per read as the read with the most 5’ alignment coordinate to a transcript among those that aligned to the correct strand and had an alignment score of no more than 4 below the best alignment score overall. We then iteratively collapsed pairs of UMIs  if the sum of the minimum Phred quality scores across non matching bases was at most 30 to account for mis matched bases that were likely due to sequencing errors. A collapsed pair was represented for future comparisons by the UMI with a larger prior number of aligned reads unless the other UMI had at least 75% as many reads aligned and had equal or better UMI quality scores for those alignments evaluated as the softmin over aligned reads of the softmax over quality scores for the UMI bases in an aligned read. post collapsing  the number of transcripts observed per gene was calculated. This value was then adjusted based on the probability of a collision the drawing at random of the same UMI multiple times  which was calculated from the number of UMIs not detected for a given gene. The full length fragmented reads without xxx’ enrichment or UMI were aligned using RSEM against a reference transcriptome  which is depleted of mtRNA  rRNA  and pseudogenes as described above. The command ‘rsem calculate expression’ was used with parameters  p 4   paired end   seed length 24 for performing the alignment. The resulting gene levels in FPKM was used to build the expression matrix. Genome build: GRCz10 dr81 Supplementary files format and content: Tab delimited text files include gene levels UMI counts or FKPM for each sample.,Embryonic cell,,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for 5’ enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,Fish of each genotype were incrossed and embryos were collected. Embryos were dechorionated and allowed to grow at 28°C in standard fish water containing methylene blue till 50% epiboly stage.,developmental stage:50% epiboly|library preparation batch:p4|genotype:Wild type TLAB,GSM2838454,GSM2838454: wt p4 S3; Danio rerio; RNA Seq,GSM2838454,,1,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for five prime enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina HiSeq 2500,,SRP123526,,,wt_p4_S3.R1.fastq.gz wt_p4_S3.R2.fastq.gz,fastq fastq,74236232.0,1325647.0,GSM2838454 r1,0:24 1:32,A:20239486;C:16647872;G:16937826;T:20349461;N:61587,24,32,,,20239486,16647872,16937826,20349461,61587,SRX3357294,SRS2656376,SRA627905,GEO,"Schier Lab, Molecular and Cellular Biology, Harvard University",2,0.77839,0.83033,0.13135,0.1391,0.84338,0.8437,0.70804,0.71279,24,32,B,B,biological fallback assumption,illumina,hiseq_era,full_length,poly_a,unknown,sc,single_cell_plate,smartseq,,United States,2017-11-02,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
44083,SRR6250367,SRX3357293,SRS2656374,SRP123526,PRJNA416939,Single cell RNAseq SMART seq2 of wild type TLAB and MZoep tz57 zebrafish embryos at 50% epiboly stage,GSE106466,Transcriptome Analysis,SMART seq2 was performed on single cells isolated from visually staged zebrafish embryos. Overall design: Samples were all sequenced in one batch. Some were generated with a five prime' UMI tagged method  and others are full length SMART seq2.,parent bioproject:PRJNA417291,pubmed:29700225,,wt p4 S2,GSM2838453,,tissue:Embryonic cell|developmental stage:50% epiboly|library preparation batch:p4|genotype:Wild type TLAB,wt p4 S2,Bases 1 8 the UMI and template switch were trimmed from the remaining reads  trimmed stretches of polyA or polyT longer than 6bp and removed any reads that were shorter than 15 bp post trimming. We identified and depleted mtRNA and rRNA reads from the remaining reads by using Bowtie2 to align all reads against a transcriptome of mtRNA and rRNA  using the parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed   no unal. Remaining reads were then aligned to a modified version of the Ensembl Zv10 release 81 zebrafish transcriptome using Bowtie2 with parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed. The transcriptome start site of each gene was padded by 100bp to increase alignment rates  since reads were generated from the 5’ end of the transcript  and not all transcripts had correct annotations for their 5’ end. Additionally  pseudogenes were removed from the transcriptome as they sometimes interfered with real genes  given our 5’ coordinate selection described below. We chose the best alignment per read as the read with the most 5’ alignment coordinate to a transcript among those that aligned to the correct strand and had an alignment score of no more than 4 below the best alignment score overall. We then iteratively collapsed pairs of UMIs  if the sum of the minimum Phred quality scores across non matching bases was at most 30 to account for mis matched bases that were likely due to sequencing errors. A collapsed pair was represented for future comparisons by the UMI with a larger prior number of aligned reads unless the other UMI had at least 75% as many reads aligned and had equal or better UMI quality scores for those alignments evaluated as the softmin over aligned reads of the softmax over quality scores for the UMI bases in an aligned read. post collapsing  the number of transcripts observed per gene was calculated. This value was then adjusted based on the probability of a collision the drawing at random of the same UMI multiple times  which was calculated from the number of UMIs not detected for a given gene. The full length fragmented reads without xxx’ enrichment or UMI were aligned using RSEM against a reference transcriptome  which is depleted of mtRNA  rRNA  and pseudogenes as described above. The command ‘rsem calculate expression’ was used with parameters  p 4   paired end   seed length 24 for performing the alignment. The resulting gene levels in FPKM was used to build the expression matrix. Genome build: GRCz10 dr81 Supplementary files format and content: Tab delimited text files include gene levels UMI counts or FKPM for each sample.,Embryonic cell,,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for 5’ enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,Fish of each genotype were incrossed and embryos were collected. Embryos were dechorionated and allowed to grow at 28°C in standard fish water containing methylene blue till 50% epiboly stage.,developmental stage:50% epiboly|library preparation batch:p4|genotype:Wild type TLAB,GSM2838453,GSM2838453: wt p4 S2; Danio rerio; RNA Seq,GSM2838453,,1,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for five prime enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina HiSeq 2500,,SRP123526,,,wt_p4_S2.R1.fastq.gz wt_p4_S2.R2.fastq.gz,fastq fastq,75103448.0,1341133.0,GSM2838453 r1,0:24 1:32,A:19864937;C:17468556;G:17687706;T:20021624;N:60625,24,32,,,19864937,17468556,17687706,20021624,60625,SRX3357293,SRS2656374,SRA627905,GEO,"Schier Lab, Molecular and Cellular Biology, Harvard University",2,0.78076,0.83253,0.11443,0.12223,0.84758,0.84774,0.71769,0.71187,24,32,B,B,biological fallback assumption,illumina,hiseq_era,full_length,poly_a,unknown,sc,single_cell_plate,smartseq,,United States,2017-11-02,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
44084,SRR6250366,SRX3357292,SRS2656375,SRP123526,PRJNA416939,Single cell RNAseq SMART seq2 of wild type TLAB and MZoep tz57 zebrafish embryos at 50% epiboly stage,GSE106466,Transcriptome Analysis,SMART seq2 was performed on single cells isolated from visually staged zebrafish embryos. Overall design: Samples were all sequenced in one batch. Some were generated with a five prime' UMI tagged method  and others are full length SMART seq2.,parent bioproject:PRJNA417291,pubmed:29700225,,wt p4 S1,GSM2838452,,tissue:Embryonic cell|developmental stage:50% epiboly|library preparation batch:p4|genotype:Wild type TLAB,wt p4 S1,Bases 1 8 the UMI and template switch were trimmed from the remaining reads  trimmed stretches of polyA or polyT longer than 6bp and removed any reads that were shorter than 15 bp post trimming. We identified and depleted mtRNA and rRNA reads from the remaining reads by using Bowtie2 to align all reads against a transcriptome of mtRNA and rRNA  using the parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed   no unal. Remaining reads were then aligned to a modified version of the Ensembl Zv10 release 81 zebrafish transcriptome using Bowtie2 with parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed. The transcriptome start site of each gene was padded by 100bp to increase alignment rates  since reads were generated from the 5’ end of the transcript  and not all transcripts had correct annotations for their 5’ end. Additionally  pseudogenes were removed from the transcriptome as they sometimes interfered with real genes  given our 5’ coordinate selection described below. We chose the best alignment per read as the read with the most 5’ alignment coordinate to a transcript among those that aligned to the correct strand and had an alignment score of no more than 4 below the best alignment score overall. We then iteratively collapsed pairs of UMIs  if the sum of the minimum Phred quality scores across non matching bases was at most 30 to account for mis matched bases that were likely due to sequencing errors. A collapsed pair was represented for future comparisons by the UMI with a larger prior number of aligned reads unless the other UMI had at least 75% as many reads aligned and had equal or better UMI quality scores for those alignments evaluated as the softmin over aligned reads of the softmax over quality scores for the UMI bases in an aligned read. post collapsing  the number of transcripts observed per gene was calculated. This value was then adjusted based on the probability of a collision the drawing at random of the same UMI multiple times  which was calculated from the number of UMIs not detected for a given gene. The full length fragmented reads without xxx’ enrichment or UMI were aligned using RSEM against a reference transcriptome  which is depleted of mtRNA  rRNA  and pseudogenes as described above. The command ‘rsem calculate expression’ was used with parameters  p 4   paired end   seed length 24 for performing the alignment. The resulting gene levels in FPKM was used to build the expression matrix. Genome build: GRCz10 dr81 Supplementary files format and content: Tab delimited text files include gene levels UMI counts or FKPM for each sample.,Embryonic cell,,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for 5’ enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,Fish of each genotype were incrossed and embryos were collected. Embryos were dechorionated and allowed to grow at 28°C in standard fish water containing methylene blue till 50% epiboly stage.,developmental stage:50% epiboly|library preparation batch:p4|genotype:Wild type TLAB,GSM2838452,GSM2838452: wt p4 S1; Danio rerio; RNA Seq,GSM2838452,,1,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for five prime enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina HiSeq 2500,,SRP123526,,,wt_p4_S1.R1.fastq.gz wt_p4_S1.R2.fastq.gz,fastq fastq,30752288.0,549148.0,GSM2838452 r1,0:24 1:32,A:8244560;C:7083728;G:7062913;T:8335315;N:25772,24,32,,,8244560,7083728,7062913,8335315,25772,SRX3357292,SRS2656375,SRA627905,GEO,"Schier Lab, Molecular and Cellular Biology, Harvard University",2,0.76119,0.82205,0.12638,0.13376,0.8857,0.88532,0.7951,0.7962,24,32,B,B,biological fallback assumption,illumina,hiseq_era,full_length,poly_a,unknown,sc,single_cell_plate,smartseq,,United States,2017-11-02,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
44085,SRR6250365,SRX3357291,SRS2656372,SRP123526,PRJNA416939,Single cell RNAseq SMART seq2 of wild type TLAB and MZoep tz57 zebrafish embryos at 50% epiboly stage,GSE106466,Transcriptome Analysis,SMART seq2 was performed on single cells isolated from visually staged zebrafish embryos. Overall design: Samples were all sequenced in one batch. Some were generated with a five prime' UMI tagged method  and others are full length SMART seq2.,parent bioproject:PRJNA417291,pubmed:29700225,,wt p4 S13,GSM2838451,,tissue:Embryonic cell|developmental stage:50% epiboly|library preparation batch:p4|genotype:Wild type TLAB,wt p4 S13,Bases 1 8 the UMI and template switch were trimmed from the remaining reads  trimmed stretches of polyA or polyT longer than 6bp and removed any reads that were shorter than 15 bp post trimming. We identified and depleted mtRNA and rRNA reads from the remaining reads by using Bowtie2 to align all reads against a transcriptome of mtRNA and rRNA  using the parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed   no unal. Remaining reads were then aligned to a modified version of the Ensembl Zv10 release 81 zebrafish transcriptome using Bowtie2 with parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed. The transcriptome start site of each gene was padded by 100bp to increase alignment rates  since reads were generated from the 5’ end of the transcript  and not all transcripts had correct annotations for their 5’ end. Additionally  pseudogenes were removed from the transcriptome as they sometimes interfered with real genes  given our 5’ coordinate selection described below. We chose the best alignment per read as the read with the most 5’ alignment coordinate to a transcript among those that aligned to the correct strand and had an alignment score of no more than 4 below the best alignment score overall. We then iteratively collapsed pairs of UMIs  if the sum of the minimum Phred quality scores across non matching bases was at most 30 to account for mis matched bases that were likely due to sequencing errors. A collapsed pair was represented for future comparisons by the UMI with a larger prior number of aligned reads unless the other UMI had at least 75% as many reads aligned and had equal or better UMI quality scores for those alignments evaluated as the softmin over aligned reads of the softmax over quality scores for the UMI bases in an aligned read. post collapsing  the number of transcripts observed per gene was calculated. This value was then adjusted based on the probability of a collision the drawing at random of the same UMI multiple times  which was calculated from the number of UMIs not detected for a given gene. The full length fragmented reads without xxx’ enrichment or UMI were aligned using RSEM against a reference transcriptome  which is depleted of mtRNA  rRNA  and pseudogenes as described above. The command ‘rsem calculate expression’ was used with parameters  p 4   paired end   seed length 24 for performing the alignment. The resulting gene levels in FPKM was used to build the expression matrix. Genome build: GRCz10 dr81 Supplementary files format and content: Tab delimited text files include gene levels UMI counts or FKPM for each sample.,Embryonic cell,,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for 5’ enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,Fish of each genotype were incrossed and embryos were collected. Embryos were dechorionated and allowed to grow at 28°C in standard fish water containing methylene blue till 50% epiboly stage.,developmental stage:50% epiboly|library preparation batch:p4|genotype:Wild type TLAB,GSM2838451,GSM2838451: wt p4 S13; Danio rerio; RNA Seq,GSM2838451,,1,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for five prime enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina HiSeq 2500,,SRP123526,,,wt_p4_S13.R1.fastq.gz wt_p4_S13.R2.fastq.gz,fastq fastq,85813392.0,1532382.0,GSM2838451 r1,0:24 1:32,A:23602233;C:19100017;G:19492573;T:23547883;N:70686,24,32,,,23602233,19100017,19492573,23547883,70686,SRX3357291,SRS2656372,SRA627905,GEO,"Schier Lab, Molecular and Cellular Biology, Harvard University",2,0.77143,0.82139,0.09157,0.09496,0.86257,0.86273,0.72727,0.72502,24,32,B,B,biological fallback assumption,illumina,hiseq_era,full_length,poly_a,unknown,sc,single_cell_plate,smartseq,,United States,2017-11-02,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
44086,SRR6250364,SRX3357290,SRS2656373,SRP123526,PRJNA416939,Single cell RNAseq SMART seq2 of wild type TLAB and MZoep tz57 zebrafish embryos at 50% epiboly stage,GSE106466,Transcriptome Analysis,SMART seq2 was performed on single cells isolated from visually staged zebrafish embryos. Overall design: Samples were all sequenced in one batch. Some were generated with a five prime' UMI tagged method  and others are full length SMART seq2.,parent bioproject:PRJNA417291,pubmed:29700225,,wt p4 S12,GSM2838450,,tissue:Embryonic cell|developmental stage:50% epiboly|library preparation batch:p4|genotype:Wild type TLAB,wt p4 S12,Bases 1 8 the UMI and template switch were trimmed from the remaining reads  trimmed stretches of polyA or polyT longer than 6bp and removed any reads that were shorter than 15 bp post trimming. We identified and depleted mtRNA and rRNA reads from the remaining reads by using Bowtie2 to align all reads against a transcriptome of mtRNA and rRNA  using the parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed   no unal. Remaining reads were then aligned to a modified version of the Ensembl Zv10 release 81 zebrafish transcriptome using Bowtie2 with parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed. The transcriptome start site of each gene was padded by 100bp to increase alignment rates  since reads were generated from the 5’ end of the transcript  and not all transcripts had correct annotations for their 5’ end. Additionally  pseudogenes were removed from the transcriptome as they sometimes interfered with real genes  given our 5’ coordinate selection described below. We chose the best alignment per read as the read with the most 5’ alignment coordinate to a transcript among those that aligned to the correct strand and had an alignment score of no more than 4 below the best alignment score overall. We then iteratively collapsed pairs of UMIs  if the sum of the minimum Phred quality scores across non matching bases was at most 30 to account for mis matched bases that were likely due to sequencing errors. A collapsed pair was represented for future comparisons by the UMI with a larger prior number of aligned reads unless the other UMI had at least 75% as many reads aligned and had equal or better UMI quality scores for those alignments evaluated as the softmin over aligned reads of the softmax over quality scores for the UMI bases in an aligned read. post collapsing  the number of transcripts observed per gene was calculated. This value was then adjusted based on the probability of a collision the drawing at random of the same UMI multiple times  which was calculated from the number of UMIs not detected for a given gene. The full length fragmented reads without xxx’ enrichment or UMI were aligned using RSEM against a reference transcriptome  which is depleted of mtRNA  rRNA  and pseudogenes as described above. The command ‘rsem calculate expression’ was used with parameters  p 4   paired end   seed length 24 for performing the alignment. The resulting gene levels in FPKM was used to build the expression matrix. Genome build: GRCz10 dr81 Supplementary files format and content: Tab delimited text files include gene levels UMI counts or FKPM for each sample.,Embryonic cell,,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for 5’ enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,Fish of each genotype were incrossed and embryos were collected. Embryos were dechorionated and allowed to grow at 28°C in standard fish water containing methylene blue till 50% epiboly stage.,developmental stage:50% epiboly|library preparation batch:p4|genotype:Wild type TLAB,GSM2838450,GSM2838450: wt p4 S12; Danio rerio; RNA Seq,GSM2838450,,1,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for five prime enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina HiSeq 2500,,SRP123526,,,wt_p4_S12.R1.fastq.gz wt_p4_S12.R2.fastq.gz,fastq fastq,18404008.0,328643.0,GSM2838450 r1,0:24 1:32,A:5125310;C:3960294;G:4054209;T:5248856;N:15339,24,32,,,5125310,3960294,4054209,5248856,15339,SRX3357290,SRS2656373,SRA627905,GEO,"Schier Lab, Molecular and Cellular Biology, Harvard University",2,0.76331,0.8125,0.0999,0.1021,0.87028,0.87022,0.3061,0.7522,24,32,B,B,biological fallback assumption,illumina,hiseq_era,full_length,poly_a,unknown,sc,single_cell_plate,smartseq,,United States,2017-11-02,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
44087,SRR6250363,SRX3357289,SRS2656371,SRP123526,PRJNA416939,Single cell RNAseq SMART seq2 of wild type TLAB and MZoep tz57 zebrafish embryos at 50% epiboly stage,GSE106466,Transcriptome Analysis,SMART seq2 was performed on single cells isolated from visually staged zebrafish embryos. Overall design: Samples were all sequenced in one batch. Some were generated with a five prime' UMI tagged method  and others are full length SMART seq2.,parent bioproject:PRJNA417291,pubmed:29700225,,wt p4 S11,GSM2838449,,tissue:Embryonic cell|developmental stage:50% epiboly|library preparation batch:p4|genotype:Wild type TLAB,wt p4 S11,Bases 1 8 the UMI and template switch were trimmed from the remaining reads  trimmed stretches of polyA or polyT longer than 6bp and removed any reads that were shorter than 15 bp post trimming. We identified and depleted mtRNA and rRNA reads from the remaining reads by using Bowtie2 to align all reads against a transcriptome of mtRNA and rRNA  using the parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed   no unal. Remaining reads were then aligned to a modified version of the Ensembl Zv10 release 81 zebrafish transcriptome using Bowtie2 with parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed. The transcriptome start site of each gene was padded by 100bp to increase alignment rates  since reads were generated from the 5’ end of the transcript  and not all transcripts had correct annotations for their 5’ end. Additionally  pseudogenes were removed from the transcriptome as they sometimes interfered with real genes  given our 5’ coordinate selection described below. We chose the best alignment per read as the read with the most 5’ alignment coordinate to a transcript among those that aligned to the correct strand and had an alignment score of no more than 4 below the best alignment score overall. We then iteratively collapsed pairs of UMIs  if the sum of the minimum Phred quality scores across non matching bases was at most 30 to account for mis matched bases that were likely due to sequencing errors. A collapsed pair was represented for future comparisons by the UMI with a larger prior number of aligned reads unless the other UMI had at least 75% as many reads aligned and had equal or better UMI quality scores for those alignments evaluated as the softmin over aligned reads of the softmax over quality scores for the UMI bases in an aligned read. post collapsing  the number of transcripts observed per gene was calculated. This value was then adjusted based on the probability of a collision the drawing at random of the same UMI multiple times  which was calculated from the number of UMIs not detected for a given gene. The full length fragmented reads without xxx’ enrichment or UMI were aligned using RSEM against a reference transcriptome  which is depleted of mtRNA  rRNA  and pseudogenes as described above. The command ‘rsem calculate expression’ was used with parameters  p 4   paired end   seed length 24 for performing the alignment. The resulting gene levels in FPKM was used to build the expression matrix. Genome build: GRCz10 dr81 Supplementary files format and content: Tab delimited text files include gene levels UMI counts or FKPM for each sample.,Embryonic cell,,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for 5’ enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,Fish of each genotype were incrossed and embryos were collected. Embryos were dechorionated and allowed to grow at 28°C in standard fish water containing methylene blue till 50% epiboly stage.,developmental stage:50% epiboly|library preparation batch:p4|genotype:Wild type TLAB,GSM2838449,GSM2838449: wt p4 S11; Danio rerio; RNA Seq,GSM2838449,,1,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for five prime enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina HiSeq 2500,,SRP123526,,,wt_p4_S11.R1.fastq.gz wt_p4_S11.R2.fastq.gz,fastq fastq,81838568.0,1461403.0,GSM2838449 r1,0:24 1:32,A:22067792;C:18616034;G:18950203;T:22137237;N:67302,24,32,,,22067792,18616034,18950203,22137237,67302,SRX3357289,SRS2656371,SRA627905,GEO,"Schier Lab, Molecular and Cellular Biology, Harvard University",2,0.77673,0.82676,0.11823,0.12552,0.8619,0.86399,0.73261,0.73128,24,32,B,B,biological fallback assumption,illumina,hiseq_era,full_length,poly_a,unknown,sc,single_cell_plate,smartseq,,United States,2017-11-02,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
44088,SRR6250362,SRX3357288,SRS2656370,SRP123526,PRJNA416939,Single cell RNAseq SMART seq2 of wild type TLAB and MZoep tz57 zebrafish embryos at 50% epiboly stage,GSE106466,Transcriptome Analysis,SMART seq2 was performed on single cells isolated from visually staged zebrafish embryos. Overall design: Samples were all sequenced in one batch. Some were generated with a five prime' UMI tagged method  and others are full length SMART seq2.,parent bioproject:PRJNA417291,pubmed:29700225,,wt p4 S10,GSM2838448,,tissue:Embryonic cell|developmental stage:50% epiboly|library preparation batch:p4|genotype:Wild type TLAB,wt p4 S10,Bases 1 8 the UMI and template switch were trimmed from the remaining reads  trimmed stretches of polyA or polyT longer than 6bp and removed any reads that were shorter than 15 bp post trimming. We identified and depleted mtRNA and rRNA reads from the remaining reads by using Bowtie2 to align all reads against a transcriptome of mtRNA and rRNA  using the parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed   no unal. Remaining reads were then aligned to a modified version of the Ensembl Zv10 release 81 zebrafish transcriptome using Bowtie2 with parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed. The transcriptome start site of each gene was padded by 100bp to increase alignment rates  since reads were generated from the 5’ end of the transcript  and not all transcripts had correct annotations for their 5’ end. Additionally  pseudogenes were removed from the transcriptome as they sometimes interfered with real genes  given our 5’ coordinate selection described below. We chose the best alignment per read as the read with the most 5’ alignment coordinate to a transcript among those that aligned to the correct strand and had an alignment score of no more than 4 below the best alignment score overall. We then iteratively collapsed pairs of UMIs  if the sum of the minimum Phred quality scores across non matching bases was at most 30 to account for mis matched bases that were likely due to sequencing errors. A collapsed pair was represented for future comparisons by the UMI with a larger prior number of aligned reads unless the other UMI had at least 75% as many reads aligned and had equal or better UMI quality scores for those alignments evaluated as the softmin over aligned reads of the softmax over quality scores for the UMI bases in an aligned read. post collapsing  the number of transcripts observed per gene was calculated. This value was then adjusted based on the probability of a collision the drawing at random of the same UMI multiple times  which was calculated from the number of UMIs not detected for a given gene. The full length fragmented reads without xxx’ enrichment or UMI were aligned using RSEM against a reference transcriptome  which is depleted of mtRNA  rRNA  and pseudogenes as described above. The command ‘rsem calculate expression’ was used with parameters  p 4   paired end   seed length 24 for performing the alignment. The resulting gene levels in FPKM was used to build the expression matrix. Genome build: GRCz10 dr81 Supplementary files format and content: Tab delimited text files include gene levels UMI counts or FKPM for each sample.,Embryonic cell,,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for 5’ enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,Fish of each genotype were incrossed and embryos were collected. Embryos were dechorionated and allowed to grow at 28°C in standard fish water containing methylene blue till 50% epiboly stage.,developmental stage:50% epiboly|library preparation batch:p4|genotype:Wild type TLAB,GSM2838448,GSM2838448: wt p4 S10; Danio rerio; RNA Seq,GSM2838448,,1,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for five prime enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina HiSeq 2500,,SRP123526,,,wt_p4_S10.R1.fastq.gz wt_p4_S10.R2.fastq.gz,fastq fastq,77205912.0,1378677.0,GSM2838448 r1,0:24 1:32,A:20905304;C:17521218;G:17939726;T:20777228;N:62436,24,32,,,20905304,17521218,17939726,20777228,62436,SRX3357288,SRS2656370,SRA627905,GEO,"Schier Lab, Molecular and Cellular Biology, Harvard University",2,0.77761,0.83011,0.12382,0.13254,0.86429,0.8647,0.78439,0.78658,24,32,B,B,biological fallback assumption,illumina,hiseq_era,full_length,poly_a,unknown,sc,single_cell_plate,smartseq,,United States,2017-11-02,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
44089,SRR6250361,SRX3357287,SRS2656369,SRP123526,PRJNA416939,Single cell RNAseq SMART seq2 of wild type TLAB and MZoep tz57 zebrafish embryos at 50% epiboly stage,GSE106466,Transcriptome Analysis,SMART seq2 was performed on single cells isolated from visually staged zebrafish embryos. Overall design: Samples were all sequenced in one batch. Some were generated with a five prime' UMI tagged method  and others are full length SMART seq2.,parent bioproject:PRJNA417291,pubmed:29700225,,wt p3 S9,GSM2838447,,tissue:Embryonic cell|developmental stage:50% epiboly|library preparation batch:p3|genotype:Wild type TLAB,wt p3 S9,Bases 1 8 the UMI and template switch were trimmed from the remaining reads  trimmed stretches of polyA or polyT longer than 6bp and removed any reads that were shorter than 15 bp post trimming. We identified and depleted mtRNA and rRNA reads from the remaining reads by using Bowtie2 to align all reads against a transcriptome of mtRNA and rRNA  using the parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed   no unal. Remaining reads were then aligned to a modified version of the Ensembl Zv10 release 81 zebrafish transcriptome using Bowtie2 with parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed. The transcriptome start site of each gene was padded by 100bp to increase alignment rates  since reads were generated from the 5’ end of the transcript  and not all transcripts had correct annotations for their 5’ end. Additionally  pseudogenes were removed from the transcriptome as they sometimes interfered with real genes  given our 5’ coordinate selection described below. We chose the best alignment per read as the read with the most 5’ alignment coordinate to a transcript among those that aligned to the correct strand and had an alignment score of no more than 4 below the best alignment score overall. We then iteratively collapsed pairs of UMIs  if the sum of the minimum Phred quality scores across non matching bases was at most 30 to account for mis matched bases that were likely due to sequencing errors. A collapsed pair was represented for future comparisons by the UMI with a larger prior number of aligned reads unless the other UMI had at least 75% as many reads aligned and had equal or better UMI quality scores for those alignments evaluated as the softmin over aligned reads of the softmax over quality scores for the UMI bases in an aligned read. post collapsing  the number of transcripts observed per gene was calculated. This value was then adjusted based on the probability of a collision the drawing at random of the same UMI multiple times  which was calculated from the number of UMIs not detected for a given gene. The full length fragmented reads without xxx’ enrichment or UMI were aligned using RSEM against a reference transcriptome  which is depleted of mtRNA  rRNA  and pseudogenes as described above. The command ‘rsem calculate expression’ was used with parameters  p 4   paired end   seed length 24 for performing the alignment. The resulting gene levels in FPKM was used to build the expression matrix. Genome build: GRCz10 dr81 Supplementary files format and content: Tab delimited text files include gene levels UMI counts or FKPM for each sample.,Embryonic cell,,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for 5’ enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,Fish of each genotype were incrossed and embryos were collected. Embryos were dechorionated and allowed to grow at 28°C in standard fish water containing methylene blue till 50% epiboly stage.,developmental stage:50% epiboly|library preparation batch:p3|genotype:Wild type TLAB,GSM2838447,GSM2838447: wt p3 S9; Danio rerio; RNA Seq,GSM2838447,,1,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for five prime enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina HiSeq 2500,,SRP123526,,,wt_p3_S9.R1.fastq.gz wt_p3_S9.R2.fastq.gz,fastq fastq,2408.0,43.0,GSM2838447 r1,0:24 1:32,A:630;C:591;G:610;T:577;N:0,24,32,,,630,591,610,577,0,SRX3357287,SRS2656369,SRA627905,GEO,"Schier Lab, Molecular and Cellular Biology, Harvard University",2,0.91429,0.36364,0.2,0.0909,0.99963,0.99995,0.68,0.66666,24,32,B,T,mate2 technical by mapping diff,illumina,hiseq_era,full_length,poly_a,unknown,sc,single_cell_plate,smartseq,,United States,2017-11-02,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
44090,SRR6250360,SRX3357286,SRS2656368,SRP123526,PRJNA416939,Single cell RNAseq SMART seq2 of wild type TLAB and MZoep tz57 zebrafish embryos at 50% epiboly stage,GSE106466,Transcriptome Analysis,SMART seq2 was performed on single cells isolated from visually staged zebrafish embryos. Overall design: Samples were all sequenced in one batch. Some were generated with a five prime' UMI tagged method  and others are full length SMART seq2.,parent bioproject:PRJNA417291,pubmed:29700225,,wt p3 S8,GSM2838446,,tissue:Embryonic cell|developmental stage:50% epiboly|library preparation batch:p3|genotype:Wild type TLAB,wt p3 S8,Bases 1 8 the UMI and template switch were trimmed from the remaining reads  trimmed stretches of polyA or polyT longer than 6bp and removed any reads that were shorter than 15 bp post trimming. We identified and depleted mtRNA and rRNA reads from the remaining reads by using Bowtie2 to align all reads against a transcriptome of mtRNA and rRNA  using the parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed   no unal. Remaining reads were then aligned to a modified version of the Ensembl Zv10 release 81 zebrafish transcriptome using Bowtie2 with parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed. The transcriptome start site of each gene was padded by 100bp to increase alignment rates  since reads were generated from the 5’ end of the transcript  and not all transcripts had correct annotations for their 5’ end. Additionally  pseudogenes were removed from the transcriptome as they sometimes interfered with real genes  given our 5’ coordinate selection described below. We chose the best alignment per read as the read with the most 5’ alignment coordinate to a transcript among those that aligned to the correct strand and had an alignment score of no more than 4 below the best alignment score overall. We then iteratively collapsed pairs of UMIs  if the sum of the minimum Phred quality scores across non matching bases was at most 30 to account for mis matched bases that were likely due to sequencing errors. A collapsed pair was represented for future comparisons by the UMI with a larger prior number of aligned reads unless the other UMI had at least 75% as many reads aligned and had equal or better UMI quality scores for those alignments evaluated as the softmin over aligned reads of the softmax over quality scores for the UMI bases in an aligned read. post collapsing  the number of transcripts observed per gene was calculated. This value was then adjusted based on the probability of a collision the drawing at random of the same UMI multiple times  which was calculated from the number of UMIs not detected for a given gene. The full length fragmented reads without xxx’ enrichment or UMI were aligned using RSEM against a reference transcriptome  which is depleted of mtRNA  rRNA  and pseudogenes as described above. The command ‘rsem calculate expression’ was used with parameters  p 4   paired end   seed length 24 for performing the alignment. The resulting gene levels in FPKM was used to build the expression matrix. Genome build: GRCz10 dr81 Supplementary files format and content: Tab delimited text files include gene levels UMI counts or FKPM for each sample.,Embryonic cell,,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for 5’ enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,Fish of each genotype were incrossed and embryos were collected. Embryos were dechorionated and allowed to grow at 28°C in standard fish water containing methylene blue till 50% epiboly stage.,developmental stage:50% epiboly|library preparation batch:p3|genotype:Wild type TLAB,GSM2838446,GSM2838446: wt p3 S8; Danio rerio; RNA Seq,GSM2838446,,1,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for five prime enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina HiSeq 2500,,SRP123526,,,wt_p3_S8.R1.fastq.gz wt_p3_S8.R2.fastq.gz,fastq fastq,1960.0,35.0,GSM2838446 r1,0:24 1:32,A:469;C:513;G:516;T:462;N:0,24,32,,,469,513,516,462,0,SRX3357286,SRS2656368,SRA627905,GEO,"Schier Lab, Molecular and Cellular Biology, Harvard University",2,0.88,0.25,0.2,0.0,0.99975,0.99995,0.82352,1.0,24,32,B,B,mate2-mate1 similar by mapping diff,illumina,hiseq_era,full_length,poly_a,unknown,sc,single_cell_plate,smartseq,,United States,2017-11-02,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
44091,SRR6250359,SRX3357284,SRS2656366,SRP123526,PRJNA416939,Single cell RNAseq SMART seq2 of wild type TLAB and MZoep tz57 zebrafish embryos at 50% epiboly stage,GSE106466,Transcriptome Analysis,SMART seq2 was performed on single cells isolated from visually staged zebrafish embryos. Overall design: Samples were all sequenced in one batch. Some were generated with a five prime' UMI tagged method  and others are full length SMART seq2.,parent bioproject:PRJNA417291,pubmed:29700225,,wt p3 S7,GSM2838445,,tissue:Embryonic cell|developmental stage:50% epiboly|library preparation batch:p3|genotype:Wild type TLAB,wt p3 S7,Bases 1 8 the UMI and template switch were trimmed from the remaining reads  trimmed stretches of polyA or polyT longer than 6bp and removed any reads that were shorter than 15 bp post trimming. We identified and depleted mtRNA and rRNA reads from the remaining reads by using Bowtie2 to align all reads against a transcriptome of mtRNA and rRNA  using the parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed   no unal. Remaining reads were then aligned to a modified version of the Ensembl Zv10 release 81 zebrafish transcriptome using Bowtie2 with parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed. The transcriptome start site of each gene was padded by 100bp to increase alignment rates  since reads were generated from the 5’ end of the transcript  and not all transcripts had correct annotations for their 5’ end. Additionally  pseudogenes were removed from the transcriptome as they sometimes interfered with real genes  given our 5’ coordinate selection described below. We chose the best alignment per read as the read with the most 5’ alignment coordinate to a transcript among those that aligned to the correct strand and had an alignment score of no more than 4 below the best alignment score overall. We then iteratively collapsed pairs of UMIs  if the sum of the minimum Phred quality scores across non matching bases was at most 30 to account for mis matched bases that were likely due to sequencing errors. A collapsed pair was represented for future comparisons by the UMI with a larger prior number of aligned reads unless the other UMI had at least 75% as many reads aligned and had equal or better UMI quality scores for those alignments evaluated as the softmin over aligned reads of the softmax over quality scores for the UMI bases in an aligned read. post collapsing  the number of transcripts observed per gene was calculated. This value was then adjusted based on the probability of a collision the drawing at random of the same UMI multiple times  which was calculated from the number of UMIs not detected for a given gene. The full length fragmented reads without xxx’ enrichment or UMI were aligned using RSEM against a reference transcriptome  which is depleted of mtRNA  rRNA  and pseudogenes as described above. The command ‘rsem calculate expression’ was used with parameters  p 4   paired end   seed length 24 for performing the alignment. The resulting gene levels in FPKM was used to build the expression matrix. Genome build: GRCz10 dr81 Supplementary files format and content: Tab delimited text files include gene levels UMI counts or FKPM for each sample.,Embryonic cell,,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for 5’ enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,Fish of each genotype were incrossed and embryos were collected. Embryos were dechorionated and allowed to grow at 28°C in standard fish water containing methylene blue till 50% epiboly stage.,developmental stage:50% epiboly|library preparation batch:p3|genotype:Wild type TLAB,GSM2838445,GSM2838445: wt p3 S7; Danio rerio; RNA Seq,GSM2838445,,1,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for five prime enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina HiSeq 2500,,SRP123526,,,wt_p3_S7.R1.fastq.gz wt_p3_S7.R2.fastq.gz,fastq fastq,81620280.0,1457505.0,GSM2838445 r1,0:24 1:32,A:21949726;C:18710660;G:19026412;T:21867186;N:66296,24,32,,,21949726,18710660,19026412,21867186,66296,SRX3357284,SRS2656366,SRA627905,GEO,"Schier Lab, Molecular and Cellular Biology, Harvard University",2,0.77869,0.82733,0.14008,0.14886,0.87267,0.8731,0.75824,0.75902,24,32,B,B,biological fallback assumption,illumina,hiseq_era,full_length,poly_a,unknown,sc,single_cell_plate,smartseq,,United States,2017-11-02,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
44092,SRR6250358,SRX3357283,SRS2656365,SRP123526,PRJNA416939,Single cell RNAseq SMART seq2 of wild type TLAB and MZoep tz57 zebrafish embryos at 50% epiboly stage,GSE106466,Transcriptome Analysis,SMART seq2 was performed on single cells isolated from visually staged zebrafish embryos. Overall design: Samples were all sequenced in one batch. Some were generated with a five prime' UMI tagged method  and others are full length SMART seq2.,parent bioproject:PRJNA417291,pubmed:29700225,,wt p3 S6,GSM2838444,,tissue:Embryonic cell|developmental stage:50% epiboly|library preparation batch:p3|genotype:Wild type TLAB,wt p3 S6,Bases 1 8 the UMI and template switch were trimmed from the remaining reads  trimmed stretches of polyA or polyT longer than 6bp and removed any reads that were shorter than 15 bp post trimming. We identified and depleted mtRNA and rRNA reads from the remaining reads by using Bowtie2 to align all reads against a transcriptome of mtRNA and rRNA  using the parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed   no unal. Remaining reads were then aligned to a modified version of the Ensembl Zv10 release 81 zebrafish transcriptome using Bowtie2 with parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed. The transcriptome start site of each gene was padded by 100bp to increase alignment rates  since reads were generated from the 5’ end of the transcript  and not all transcripts had correct annotations for their 5’ end. Additionally  pseudogenes were removed from the transcriptome as they sometimes interfered with real genes  given our 5’ coordinate selection described below. We chose the best alignment per read as the read with the most 5’ alignment coordinate to a transcript among those that aligned to the correct strand and had an alignment score of no more than 4 below the best alignment score overall. We then iteratively collapsed pairs of UMIs  if the sum of the minimum Phred quality scores across non matching bases was at most 30 to account for mis matched bases that were likely due to sequencing errors. A collapsed pair was represented for future comparisons by the UMI with a larger prior number of aligned reads unless the other UMI had at least 75% as many reads aligned and had equal or better UMI quality scores for those alignments evaluated as the softmin over aligned reads of the softmax over quality scores for the UMI bases in an aligned read. post collapsing  the number of transcripts observed per gene was calculated. This value was then adjusted based on the probability of a collision the drawing at random of the same UMI multiple times  which was calculated from the number of UMIs not detected for a given gene. The full length fragmented reads without xxx’ enrichment or UMI were aligned using RSEM against a reference transcriptome  which is depleted of mtRNA  rRNA  and pseudogenes as described above. The command ‘rsem calculate expression’ was used with parameters  p 4   paired end   seed length 24 for performing the alignment. The resulting gene levels in FPKM was used to build the expression matrix. Genome build: GRCz10 dr81 Supplementary files format and content: Tab delimited text files include gene levels UMI counts or FKPM for each sample.,Embryonic cell,,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for 5’ enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,Fish of each genotype were incrossed and embryos were collected. Embryos were dechorionated and allowed to grow at 28°C in standard fish water containing methylene blue till 50% epiboly stage.,developmental stage:50% epiboly|library preparation batch:p3|genotype:Wild type TLAB,GSM2838444,GSM2838444: wt p3 S6; Danio rerio; RNA Seq,GSM2838444,,1,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for five prime enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina HiSeq 2500,,SRP123526,,,wt_p3_S6.R1.fastq.gz wt_p3_S6.R2.fastq.gz,fastq fastq,84061488.0,1501098.0,GSM2838444 r1,0:24 1:32,A:22377525;C:19468302;G:19754703;T:22392074;N:68884,24,32,,,22377525,19468302,19754703,22392074,68884,SRX3357283,SRS2656365,SRA627905,GEO,"Schier Lab, Molecular and Cellular Biology, Harvard University",2,0.77689,0.82863,0.14263,0.15362,0.87349,0.87393,0.75351,0.75807,24,32,B,B,biological fallback assumption,illumina,hiseq_era,full_length,poly_a,unknown,sc,single_cell_plate,smartseq,,United States,2017-11-02,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
44093,SRR6250357,SRX3357282,SRS2656364,SRP123526,PRJNA416939,Single cell RNAseq SMART seq2 of wild type TLAB and MZoep tz57 zebrafish embryos at 50% epiboly stage,GSE106466,Transcriptome Analysis,SMART seq2 was performed on single cells isolated from visually staged zebrafish embryos. Overall design: Samples were all sequenced in one batch. Some were generated with a five prime' UMI tagged method  and others are full length SMART seq2.,parent bioproject:PRJNA417291,pubmed:29700225,,wt p3 S5,GSM2838443,,tissue:Embryonic cell|developmental stage:50% epiboly|library preparation batch:p3|genotype:Wild type TLAB,wt p3 S5,Bases 1 8 the UMI and template switch were trimmed from the remaining reads  trimmed stretches of polyA or polyT longer than 6bp and removed any reads that were shorter than 15 bp post trimming. We identified and depleted mtRNA and rRNA reads from the remaining reads by using Bowtie2 to align all reads against a transcriptome of mtRNA and rRNA  using the parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed   no unal. Remaining reads were then aligned to a modified version of the Ensembl Zv10 release 81 zebrafish transcriptome using Bowtie2 with parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed. The transcriptome start site of each gene was padded by 100bp to increase alignment rates  since reads were generated from the 5’ end of the transcript  and not all transcripts had correct annotations for their 5’ end. Additionally  pseudogenes were removed from the transcriptome as they sometimes interfered with real genes  given our 5’ coordinate selection described below. We chose the best alignment per read as the read with the most 5’ alignment coordinate to a transcript among those that aligned to the correct strand and had an alignment score of no more than 4 below the best alignment score overall. We then iteratively collapsed pairs of UMIs  if the sum of the minimum Phred quality scores across non matching bases was at most 30 to account for mis matched bases that were likely due to sequencing errors. A collapsed pair was represented for future comparisons by the UMI with a larger prior number of aligned reads unless the other UMI had at least 75% as many reads aligned and had equal or better UMI quality scores for those alignments evaluated as the softmin over aligned reads of the softmax over quality scores for the UMI bases in an aligned read. post collapsing  the number of transcripts observed per gene was calculated. This value was then adjusted based on the probability of a collision the drawing at random of the same UMI multiple times  which was calculated from the number of UMIs not detected for a given gene. The full length fragmented reads without xxx’ enrichment or UMI were aligned using RSEM against a reference transcriptome  which is depleted of mtRNA  rRNA  and pseudogenes as described above. The command ‘rsem calculate expression’ was used with parameters  p 4   paired end   seed length 24 for performing the alignment. The resulting gene levels in FPKM was used to build the expression matrix. Genome build: GRCz10 dr81 Supplementary files format and content: Tab delimited text files include gene levels UMI counts or FKPM for each sample.,Embryonic cell,,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for 5’ enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,Fish of each genotype were incrossed and embryos were collected. Embryos were dechorionated and allowed to grow at 28°C in standard fish water containing methylene blue till 50% epiboly stage.,developmental stage:50% epiboly|library preparation batch:p3|genotype:Wild type TLAB,GSM2838443,GSM2838443: wt p3 S5; Danio rerio; RNA Seq,GSM2838443,,1,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for five prime enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina HiSeq 2500,,SRP123526,,,wt_p3_S5.R1.fastq.gz wt_p3_S5.R2.fastq.gz,fastq fastq,74747288.0,1334773.0,GSM2838443 r1,0:24 1:32,A:20321491;C:16963983;G:17283545;T:20116200;N:62069,24,32,,,20321491,16963983,17283545,20116200,62069,SRX3357282,SRS2656364,SRA627905,GEO,"Schier Lab, Molecular and Cellular Biology, Harvard University",2,0.77498,0.82709,0.10755,0.11552,0.89037,0.89351,0.76636,0.27591,24,32,B,B,biological fallback assumption,illumina,hiseq_era,full_length,poly_a,unknown,sc,single_cell_plate,smartseq,,United States,2017-11-02,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
44094,SRR6250356,SRX3357281,SRS2656363,SRP123526,PRJNA416939,Single cell RNAseq SMART seq2 of wild type TLAB and MZoep tz57 zebrafish embryos at 50% epiboly stage,GSE106466,Transcriptome Analysis,SMART seq2 was performed on single cells isolated from visually staged zebrafish embryos. Overall design: Samples were all sequenced in one batch. Some were generated with a five prime' UMI tagged method  and others are full length SMART seq2.,parent bioproject:PRJNA417291,pubmed:29700225,,wt p3 S4,GSM2838442,,tissue:Embryonic cell|developmental stage:50% epiboly|library preparation batch:p3|genotype:Wild type TLAB,wt p3 S4,Bases 1 8 the UMI and template switch were trimmed from the remaining reads  trimmed stretches of polyA or polyT longer than 6bp and removed any reads that were shorter than 15 bp post trimming. We identified and depleted mtRNA and rRNA reads from the remaining reads by using Bowtie2 to align all reads against a transcriptome of mtRNA and rRNA  using the parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed   no unal. Remaining reads were then aligned to a modified version of the Ensembl Zv10 release 81 zebrafish transcriptome using Bowtie2 with parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed. The transcriptome start site of each gene was padded by 100bp to increase alignment rates  since reads were generated from the 5’ end of the transcript  and not all transcripts had correct annotations for their 5’ end. Additionally  pseudogenes were removed from the transcriptome as they sometimes interfered with real genes  given our 5’ coordinate selection described below. We chose the best alignment per read as the read with the most 5’ alignment coordinate to a transcript among those that aligned to the correct strand and had an alignment score of no more than 4 below the best alignment score overall. We then iteratively collapsed pairs of UMIs  if the sum of the minimum Phred quality scores across non matching bases was at most 30 to account for mis matched bases that were likely due to sequencing errors. A collapsed pair was represented for future comparisons by the UMI with a larger prior number of aligned reads unless the other UMI had at least 75% as many reads aligned and had equal or better UMI quality scores for those alignments evaluated as the softmin over aligned reads of the softmax over quality scores for the UMI bases in an aligned read. post collapsing  the number of transcripts observed per gene was calculated. This value was then adjusted based on the probability of a collision the drawing at random of the same UMI multiple times  which was calculated from the number of UMIs not detected for a given gene. The full length fragmented reads without xxx’ enrichment or UMI were aligned using RSEM against a reference transcriptome  which is depleted of mtRNA  rRNA  and pseudogenes as described above. The command ‘rsem calculate expression’ was used with parameters  p 4   paired end   seed length 24 for performing the alignment. The resulting gene levels in FPKM was used to build the expression matrix. Genome build: GRCz10 dr81 Supplementary files format and content: Tab delimited text files include gene levels UMI counts or FKPM for each sample.,Embryonic cell,,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for 5’ enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,Fish of each genotype were incrossed and embryos were collected. Embryos were dechorionated and allowed to grow at 28°C in standard fish water containing methylene blue till 50% epiboly stage.,developmental stage:50% epiboly|library preparation batch:p3|genotype:Wild type TLAB,GSM2838442,GSM2838442: wt p3 S4; Danio rerio; RNA Seq,GSM2838442,,1,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for five prime enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina HiSeq 2500,,SRP123526,,,wt_p3_S4.R1.fastq.gz wt_p3_S4.R2.fastq.gz,fastq fastq,39749248.0,709808.0,GSM2838442 r1,0:24 1:32,A:10051657;C:9639959;G:9805405;T:10220223;N:32004,24,32,,,10051657,9639959,9805405,10220223,32004,SRX3357281,SRS2656363,SRA627905,GEO,"Schier Lab, Molecular and Cellular Biology, Harvard University",2,0.76312,0.82257,0.18031,0.19838,0.91256,0.91441,0.69462,0.6963,24,32,B,B,biological fallback assumption,illumina,hiseq_era,full_length,poly_a,unknown,sc,single_cell_plate,smartseq,,United States,2017-11-02,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
44095,SRR6250355,SRX3357280,SRS2656362,SRP123526,PRJNA416939,Single cell RNAseq SMART seq2 of wild type TLAB and MZoep tz57 zebrafish embryos at 50% epiboly stage,GSE106466,Transcriptome Analysis,SMART seq2 was performed on single cells isolated from visually staged zebrafish embryos. Overall design: Samples were all sequenced in one batch. Some were generated with a five prime' UMI tagged method  and others are full length SMART seq2.,parent bioproject:PRJNA417291,pubmed:29700225,,wt p3 S3,GSM2838441,,tissue:Embryonic cell|developmental stage:50% epiboly|library preparation batch:p3|genotype:Wild type TLAB,wt p3 S3,Bases 1 8 the UMI and template switch were trimmed from the remaining reads  trimmed stretches of polyA or polyT longer than 6bp and removed any reads that were shorter than 15 bp post trimming. We identified and depleted mtRNA and rRNA reads from the remaining reads by using Bowtie2 to align all reads against a transcriptome of mtRNA and rRNA  using the parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed   no unal. Remaining reads were then aligned to a modified version of the Ensembl Zv10 release 81 zebrafish transcriptome using Bowtie2 with parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed. The transcriptome start site of each gene was padded by 100bp to increase alignment rates  since reads were generated from the 5’ end of the transcript  and not all transcripts had correct annotations for their 5’ end. Additionally  pseudogenes were removed from the transcriptome as they sometimes interfered with real genes  given our 5’ coordinate selection described below. We chose the best alignment per read as the read with the most 5’ alignment coordinate to a transcript among those that aligned to the correct strand and had an alignment score of no more than 4 below the best alignment score overall. We then iteratively collapsed pairs of UMIs  if the sum of the minimum Phred quality scores across non matching bases was at most 30 to account for mis matched bases that were likely due to sequencing errors. A collapsed pair was represented for future comparisons by the UMI with a larger prior number of aligned reads unless the other UMI had at least 75% as many reads aligned and had equal or better UMI quality scores for those alignments evaluated as the softmin over aligned reads of the softmax over quality scores for the UMI bases in an aligned read. post collapsing  the number of transcripts observed per gene was calculated. This value was then adjusted based on the probability of a collision the drawing at random of the same UMI multiple times  which was calculated from the number of UMIs not detected for a given gene. The full length fragmented reads without xxx’ enrichment or UMI were aligned using RSEM against a reference transcriptome  which is depleted of mtRNA  rRNA  and pseudogenes as described above. The command ‘rsem calculate expression’ was used with parameters  p 4   paired end   seed length 24 for performing the alignment. The resulting gene levels in FPKM was used to build the expression matrix. Genome build: GRCz10 dr81 Supplementary files format and content: Tab delimited text files include gene levels UMI counts or FKPM for each sample.,Embryonic cell,,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for 5’ enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,Fish of each genotype were incrossed and embryos were collected. Embryos were dechorionated and allowed to grow at 28°C in standard fish water containing methylene blue till 50% epiboly stage.,developmental stage:50% epiboly|library preparation batch:p3|genotype:Wild type TLAB,GSM2838441,GSM2838441: wt p3 S3; Danio rerio; RNA Seq,GSM2838441,,1,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for five prime enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina HiSeq 2500,,SRP123526,,,wt_p3_S3.R1.fastq.gz wt_p3_S3.R2.fastq.gz,fastq fastq,82460168.0,1472503.0,GSM2838441 r1,0:24 1:32,A:22042758;C:18983786;G:19250852;T:22115614;N:67158,24,32,,,22042758,18983786,19250852,22115614,67158,SRX3357280,SRS2656362,SRA627905,GEO,"Schier Lab, Molecular and Cellular Biology, Harvard University",2,0.77578,0.82888,0.146,0.15612,0.86036,0.85975,0.72236,0.72759,24,32,B,B,biological fallback assumption,illumina,hiseq_era,full_length,poly_a,unknown,sc,single_cell_plate,smartseq,,United States,2017-11-02,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
44096,SRR6250354,SRX3357279,SRS2656360,SRP123526,PRJNA416939,Single cell RNAseq SMART seq2 of wild type TLAB and MZoep tz57 zebrafish embryos at 50% epiboly stage,GSE106466,Transcriptome Analysis,SMART seq2 was performed on single cells isolated from visually staged zebrafish embryos. Overall design: Samples were all sequenced in one batch. Some were generated with a five prime' UMI tagged method  and others are full length SMART seq2.,parent bioproject:PRJNA417291,pubmed:29700225,,wt p3 S2,GSM2838440,,tissue:Embryonic cell|developmental stage:50% epiboly|library preparation batch:p3|genotype:Wild type TLAB,wt p3 S2,Bases 1 8 the UMI and template switch were trimmed from the remaining reads  trimmed stretches of polyA or polyT longer than 6bp and removed any reads that were shorter than 15 bp post trimming. We identified and depleted mtRNA and rRNA reads from the remaining reads by using Bowtie2 to align all reads against a transcriptome of mtRNA and rRNA  using the parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed   no unal. Remaining reads were then aligned to a modified version of the Ensembl Zv10 release 81 zebrafish transcriptome using Bowtie2 with parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed. The transcriptome start site of each gene was padded by 100bp to increase alignment rates  since reads were generated from the 5’ end of the transcript  and not all transcripts had correct annotations for their 5’ end. Additionally  pseudogenes were removed from the transcriptome as they sometimes interfered with real genes  given our 5’ coordinate selection described below. We chose the best alignment per read as the read with the most 5’ alignment coordinate to a transcript among those that aligned to the correct strand and had an alignment score of no more than 4 below the best alignment score overall. We then iteratively collapsed pairs of UMIs  if the sum of the minimum Phred quality scores across non matching bases was at most 30 to account for mis matched bases that were likely due to sequencing errors. A collapsed pair was represented for future comparisons by the UMI with a larger prior number of aligned reads unless the other UMI had at least 75% as many reads aligned and had equal or better UMI quality scores for those alignments evaluated as the softmin over aligned reads of the softmax over quality scores for the UMI bases in an aligned read. post collapsing  the number of transcripts observed per gene was calculated. This value was then adjusted based on the probability of a collision the drawing at random of the same UMI multiple times  which was calculated from the number of UMIs not detected for a given gene. The full length fragmented reads without xxx’ enrichment or UMI were aligned using RSEM against a reference transcriptome  which is depleted of mtRNA  rRNA  and pseudogenes as described above. The command ‘rsem calculate expression’ was used with parameters  p 4   paired end   seed length 24 for performing the alignment. The resulting gene levels in FPKM was used to build the expression matrix. Genome build: GRCz10 dr81 Supplementary files format and content: Tab delimited text files include gene levels UMI counts or FKPM for each sample.,Embryonic cell,,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for 5’ enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,Fish of each genotype were incrossed and embryos were collected. Embryos were dechorionated and allowed to grow at 28°C in standard fish water containing methylene blue till 50% epiboly stage.,developmental stage:50% epiboly|library preparation batch:p3|genotype:Wild type TLAB,GSM2838440,GSM2838440: wt p3 S2; Danio rerio; RNA Seq,GSM2838440,,1,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for five prime enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina HiSeq 2500,,SRP123526,,,wt_p3_S2.R1.fastq.gz wt_p3_S2.R2.fastq.gz,fastq fastq,76462176.0,1365396.0,GSM2838440 r1,0:24 1:32,A:20555351;C:17529474;G:17780648;T:20532935;N:63768,24,32,,,20555351,17529474,17780648,20532935,63768,SRX3357279,SRS2656360,SRA627905,GEO,"Schier Lab, Molecular and Cellular Biology, Harvard University",2,0.77938,0.8309,0.14163,0.15224,0.87026,0.87107,0.75973,0.76003,24,32,B,B,biological fallback assumption,illumina,hiseq_era,full_length,poly_a,unknown,sc,single_cell_plate,smartseq,,United States,2017-11-02,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
44097,SRR6250353,SRX3357278,SRS2656361,SRP123526,PRJNA416939,Single cell RNAseq SMART seq2 of wild type TLAB and MZoep tz57 zebrafish embryos at 50% epiboly stage,GSE106466,Transcriptome Analysis,SMART seq2 was performed on single cells isolated from visually staged zebrafish embryos. Overall design: Samples were all sequenced in one batch. Some were generated with a five prime' UMI tagged method  and others are full length SMART seq2.,parent bioproject:PRJNA417291,pubmed:29700225,,wt p3 S22,GSM2838439,,tissue:Embryonic cell|developmental stage:50% epiboly|library preparation batch:p3|genotype:Wild type TLAB,wt p3 S22,Bases 1 8 the UMI and template switch were trimmed from the remaining reads  trimmed stretches of polyA or polyT longer than 6bp and removed any reads that were shorter than 15 bp post trimming. We identified and depleted mtRNA and rRNA reads from the remaining reads by using Bowtie2 to align all reads against a transcriptome of mtRNA and rRNA  using the parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed   no unal. Remaining reads were then aligned to a modified version of the Ensembl Zv10 release 81 zebrafish transcriptome using Bowtie2 with parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed. The transcriptome start site of each gene was padded by 100bp to increase alignment rates  since reads were generated from the 5’ end of the transcript  and not all transcripts had correct annotations for their 5’ end. Additionally  pseudogenes were removed from the transcriptome as they sometimes interfered with real genes  given our 5’ coordinate selection described below. We chose the best alignment per read as the read with the most 5’ alignment coordinate to a transcript among those that aligned to the correct strand and had an alignment score of no more than 4 below the best alignment score overall. We then iteratively collapsed pairs of UMIs  if the sum of the minimum Phred quality scores across non matching bases was at most 30 to account for mis matched bases that were likely due to sequencing errors. A collapsed pair was represented for future comparisons by the UMI with a larger prior number of aligned reads unless the other UMI had at least 75% as many reads aligned and had equal or better UMI quality scores for those alignments evaluated as the softmin over aligned reads of the softmax over quality scores for the UMI bases in an aligned read. post collapsing  the number of transcripts observed per gene was calculated. This value was then adjusted based on the probability of a collision the drawing at random of the same UMI multiple times  which was calculated from the number of UMIs not detected for a given gene. The full length fragmented reads without xxx’ enrichment or UMI were aligned using RSEM against a reference transcriptome  which is depleted of mtRNA  rRNA  and pseudogenes as described above. The command ‘rsem calculate expression’ was used with parameters  p 4   paired end   seed length 24 for performing the alignment. The resulting gene levels in FPKM was used to build the expression matrix. Genome build: GRCz10 dr81 Supplementary files format and content: Tab delimited text files include gene levels UMI counts or FKPM for each sample.,Embryonic cell,,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for 5’ enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,Fish of each genotype were incrossed and embryos were collected. Embryos were dechorionated and allowed to grow at 28°C in standard fish water containing methylene blue till 50% epiboly stage.,developmental stage:50% epiboly|library preparation batch:p3|genotype:Wild type TLAB,GSM2838439,GSM2838439: wt p3 S22; Danio rerio; RNA Seq,GSM2838439,,1,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for five prime enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina HiSeq 2500,,SRP123526,,,wt_p3_S22.R1.fastq.gz wt_p3_S22.R2.fastq.gz,fastq fastq,2036272.0,36362.0,GSM2838439 r1,0:24 1:32,A:534051;C:482377;G:483159;T:535283;N:1402,24,32,,,534051,482377,483159,535283,1402,SRX3357278,SRS2656361,SRA627905,GEO,"Schier Lab, Molecular and Cellular Biology, Harvard University",2,0.66771,0.73881,0.15713,0.17328,0.96106,0.9598,0.80537,0.79672,24,32,B,B,biological fallback assumption,illumina,hiseq_era,full_length,poly_a,unknown,sc,single_cell_plate,smartseq,,United States,2017-11-02,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
44098,SRR6250352,SRX3357277,SRS2656359,SRP123526,PRJNA416939,Single cell RNAseq SMART seq2 of wild type TLAB and MZoep tz57 zebrafish embryos at 50% epiboly stage,GSE106466,Transcriptome Analysis,SMART seq2 was performed on single cells isolated from visually staged zebrafish embryos. Overall design: Samples were all sequenced in one batch. Some were generated with a five prime' UMI tagged method  and others are full length SMART seq2.,parent bioproject:PRJNA417291,pubmed:29700225,,wt p3 S21,GSM2838438,,tissue:Embryonic cell|developmental stage:50% epiboly|library preparation batch:p3|genotype:Wild type TLAB,wt p3 S21,Bases 1 8 the UMI and template switch were trimmed from the remaining reads  trimmed stretches of polyA or polyT longer than 6bp and removed any reads that were shorter than 15 bp post trimming. We identified and depleted mtRNA and rRNA reads from the remaining reads by using Bowtie2 to align all reads against a transcriptome of mtRNA and rRNA  using the parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed   no unal. Remaining reads were then aligned to a modified version of the Ensembl Zv10 release 81 zebrafish transcriptome using Bowtie2 with parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed. The transcriptome start site of each gene was padded by 100bp to increase alignment rates  since reads were generated from the 5’ end of the transcript  and not all transcripts had correct annotations for their 5’ end. Additionally  pseudogenes were removed from the transcriptome as they sometimes interfered with real genes  given our 5’ coordinate selection described below. We chose the best alignment per read as the read with the most 5’ alignment coordinate to a transcript among those that aligned to the correct strand and had an alignment score of no more than 4 below the best alignment score overall. We then iteratively collapsed pairs of UMIs  if the sum of the minimum Phred quality scores across non matching bases was at most 30 to account for mis matched bases that were likely due to sequencing errors. A collapsed pair was represented for future comparisons by the UMI with a larger prior number of aligned reads unless the other UMI had at least 75% as many reads aligned and had equal or better UMI quality scores for those alignments evaluated as the softmin over aligned reads of the softmax over quality scores for the UMI bases in an aligned read. post collapsing  the number of transcripts observed per gene was calculated. This value was then adjusted based on the probability of a collision the drawing at random of the same UMI multiple times  which was calculated from the number of UMIs not detected for a given gene. The full length fragmented reads without xxx’ enrichment or UMI were aligned using RSEM against a reference transcriptome  which is depleted of mtRNA  rRNA  and pseudogenes as described above. The command ‘rsem calculate expression’ was used with parameters  p 4   paired end   seed length 24 for performing the alignment. The resulting gene levels in FPKM was used to build the expression matrix. Genome build: GRCz10 dr81 Supplementary files format and content: Tab delimited text files include gene levels UMI counts or FKPM for each sample.,Embryonic cell,,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for 5’ enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,Fish of each genotype were incrossed and embryos were collected. Embryos were dechorionated and allowed to grow at 28°C in standard fish water containing methylene blue till 50% epiboly stage.,developmental stage:50% epiboly|library preparation batch:p3|genotype:Wild type TLAB,GSM2838438,GSM2838438: wt p3 S21; Danio rerio; RNA Seq,GSM2838438,,1,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for five prime enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina HiSeq 2500,,SRP123526,,,wt_p3_S21.R1.fastq.gz wt_p3_S21.R2.fastq.gz,fastq fastq,70786688.0,1264048.0,GSM2838438 r1,0:24 1:32,A:19153365;C:16090791;G:16426864;T:19057852;N:57816,24,32,,,19153365,16090791,16426864,19057852,57816,SRX3357277,SRS2656359,SRA627905,GEO,"Schier Lab, Molecular and Cellular Biology, Harvard University",2,0.77531,0.82953,0.13952,0.15031,0.89173,0.89232,0.78154,0.7875,24,32,B,B,biological fallback assumption,illumina,hiseq_era,full_length,poly_a,unknown,sc,single_cell_plate,smartseq,,United States,2017-11-02,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
44099,SRR6250351,SRX3357276,SRS2656358,SRP123526,PRJNA416939,Single cell RNAseq SMART seq2 of wild type TLAB and MZoep tz57 zebrafish embryos at 50% epiboly stage,GSE106466,Transcriptome Analysis,SMART seq2 was performed on single cells isolated from visually staged zebrafish embryos. Overall design: Samples were all sequenced in one batch. Some were generated with a five prime' UMI tagged method  and others are full length SMART seq2.,parent bioproject:PRJNA417291,pubmed:29700225,,wt p3 S20,GSM2838437,,tissue:Embryonic cell|developmental stage:50% epiboly|library preparation batch:p3|genotype:Wild type TLAB,wt p3 S20,Bases 1 8 the UMI and template switch were trimmed from the remaining reads  trimmed stretches of polyA or polyT longer than 6bp and removed any reads that were shorter than 15 bp post trimming. We identified and depleted mtRNA and rRNA reads from the remaining reads by using Bowtie2 to align all reads against a transcriptome of mtRNA and rRNA  using the parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed   no unal. Remaining reads were then aligned to a modified version of the Ensembl Zv10 release 81 zebrafish transcriptome using Bowtie2 with parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed. The transcriptome start site of each gene was padded by 100bp to increase alignment rates  since reads were generated from the 5’ end of the transcript  and not all transcripts had correct annotations for their 5’ end. Additionally  pseudogenes were removed from the transcriptome as they sometimes interfered with real genes  given our 5’ coordinate selection described below. We chose the best alignment per read as the read with the most 5’ alignment coordinate to a transcript among those that aligned to the correct strand and had an alignment score of no more than 4 below the best alignment score overall. We then iteratively collapsed pairs of UMIs  if the sum of the minimum Phred quality scores across non matching bases was at most 30 to account for mis matched bases that were likely due to sequencing errors. A collapsed pair was represented for future comparisons by the UMI with a larger prior number of aligned reads unless the other UMI had at least 75% as many reads aligned and had equal or better UMI quality scores for those alignments evaluated as the softmin over aligned reads of the softmax over quality scores for the UMI bases in an aligned read. post collapsing  the number of transcripts observed per gene was calculated. This value was then adjusted based on the probability of a collision the drawing at random of the same UMI multiple times  which was calculated from the number of UMIs not detected for a given gene. The full length fragmented reads without xxx’ enrichment or UMI were aligned using RSEM against a reference transcriptome  which is depleted of mtRNA  rRNA  and pseudogenes as described above. The command ‘rsem calculate expression’ was used with parameters  p 4   paired end   seed length 24 for performing the alignment. The resulting gene levels in FPKM was used to build the expression matrix. Genome build: GRCz10 dr81 Supplementary files format and content: Tab delimited text files include gene levels UMI counts or FKPM for each sample.,Embryonic cell,,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for 5’ enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,Fish of each genotype were incrossed and embryos were collected. Embryos were dechorionated and allowed to grow at 28°C in standard fish water containing methylene blue till 50% epiboly stage.,developmental stage:50% epiboly|library preparation batch:p3|genotype:Wild type TLAB,GSM2838437,GSM2838437: wt p3 S20; Danio rerio; RNA Seq,GSM2838437,,1,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for five prime enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina HiSeq 2500,,SRP123526,,,wt_p3_S20.R1.fastq.gz wt_p3_S20.R2.fastq.gz,fastq fastq,62717480.0,1119955.0,GSM2838437 r1,0:24 1:32,A:16708793;C:14490527;G:14735590;T:16732752;N:49818,24,32,,,16708793,14490527,14735590,16732752,49818,SRX3357276,SRS2656358,SRA627905,GEO,"Schier Lab, Molecular and Cellular Biology, Harvard University",2,0.7819,0.82495,0.17803,0.18865,0.86324,0.86302,0.69854,0.69264,24,32,B,B,biological fallback assumption,illumina,hiseq_era,full_length,poly_a,unknown,sc,single_cell_plate,smartseq,,United States,2017-11-02,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
44100,SRR6250350,SRX3357275,SRS2656357,SRP123526,PRJNA416939,Single cell RNAseq SMART seq2 of wild type TLAB and MZoep tz57 zebrafish embryos at 50% epiboly stage,GSE106466,Transcriptome Analysis,SMART seq2 was performed on single cells isolated from visually staged zebrafish embryos. Overall design: Samples were all sequenced in one batch. Some were generated with a five prime' UMI tagged method  and others are full length SMART seq2.,parent bioproject:PRJNA417291,pubmed:29700225,,wt p3 S1,GSM2838436,,tissue:Embryonic cell|developmental stage:50% epiboly|library preparation batch:p3|genotype:Wild type TLAB,wt p3 S1,Bases 1 8 the UMI and template switch were trimmed from the remaining reads  trimmed stretches of polyA or polyT longer than 6bp and removed any reads that were shorter than 15 bp post trimming. We identified and depleted mtRNA and rRNA reads from the remaining reads by using Bowtie2 to align all reads against a transcriptome of mtRNA and rRNA  using the parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed   no unal. Remaining reads were then aligned to a modified version of the Ensembl Zv10 release 81 zebrafish transcriptome using Bowtie2 with parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed. The transcriptome start site of each gene was padded by 100bp to increase alignment rates  since reads were generated from the 5’ end of the transcript  and not all transcripts had correct annotations for their 5’ end. Additionally  pseudogenes were removed from the transcriptome as they sometimes interfered with real genes  given our 5’ coordinate selection described below. We chose the best alignment per read as the read with the most 5’ alignment coordinate to a transcript among those that aligned to the correct strand and had an alignment score of no more than 4 below the best alignment score overall. We then iteratively collapsed pairs of UMIs  if the sum of the minimum Phred quality scores across non matching bases was at most 30 to account for mis matched bases that were likely due to sequencing errors. A collapsed pair was represented for future comparisons by the UMI with a larger prior number of aligned reads unless the other UMI had at least 75% as many reads aligned and had equal or better UMI quality scores for those alignments evaluated as the softmin over aligned reads of the softmax over quality scores for the UMI bases in an aligned read. post collapsing  the number of transcripts observed per gene was calculated. This value was then adjusted based on the probability of a collision the drawing at random of the same UMI multiple times  which was calculated from the number of UMIs not detected for a given gene. The full length fragmented reads without xxx’ enrichment or UMI were aligned using RSEM against a reference transcriptome  which is depleted of mtRNA  rRNA  and pseudogenes as described above. The command ‘rsem calculate expression’ was used with parameters  p 4   paired end   seed length 24 for performing the alignment. The resulting gene levels in FPKM was used to build the expression matrix. Genome build: GRCz10 dr81 Supplementary files format and content: Tab delimited text files include gene levels UMI counts or FKPM for each sample.,Embryonic cell,,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for 5’ enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,Fish of each genotype were incrossed and embryos were collected. Embryos were dechorionated and allowed to grow at 28°C in standard fish water containing methylene blue till 50% epiboly stage.,developmental stage:50% epiboly|library preparation batch:p3|genotype:Wild type TLAB,GSM2838436,GSM2838436: wt p3 S1; Danio rerio; RNA Seq,GSM2838436,,1,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for five prime enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina HiSeq 2500,,SRP123526,,,wt_p3_S1.R1.fastq.gz wt_p3_S1.R2.fastq.gz,fastq fastq,61764192.0,1102932.0,GSM2838436 r1,0:24 1:32,A:16403874;C:14300360;G:14565962;T:16444107;N:49889,24,32,,,16403874,14300360,14565962,16444107,49889,SRX3357275,SRS2656357,SRA627905,GEO,"Schier Lab, Molecular and Cellular Biology, Harvard University",2,0.7857,0.83481,0.16212,0.17189,0.85715,0.85642,0.70502,0.71291,24,32,B,B,biological fallback assumption,illumina,hiseq_era,full_length,poly_a,unknown,sc,single_cell_plate,smartseq,,United States,2017-11-02,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
44101,SRR6250349,SRX3357274,SRS2656356,SRP123526,PRJNA416939,Single cell RNAseq SMART seq2 of wild type TLAB and MZoep tz57 zebrafish embryos at 50% epiboly stage,GSE106466,Transcriptome Analysis,SMART seq2 was performed on single cells isolated from visually staged zebrafish embryos. Overall design: Samples were all sequenced in one batch. Some were generated with a five prime' UMI tagged method  and others are full length SMART seq2.,parent bioproject:PRJNA417291,pubmed:29700225,,wt p3 S19,GSM2838435,,tissue:Embryonic cell|developmental stage:50% epiboly|library preparation batch:p3|genotype:Wild type TLAB,wt p3 S19,Bases 1 8 the UMI and template switch were trimmed from the remaining reads  trimmed stretches of polyA or polyT longer than 6bp and removed any reads that were shorter than 15 bp post trimming. We identified and depleted mtRNA and rRNA reads from the remaining reads by using Bowtie2 to align all reads against a transcriptome of mtRNA and rRNA  using the parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed   no unal. Remaining reads were then aligned to a modified version of the Ensembl Zv10 release 81 zebrafish transcriptome using Bowtie2 with parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed. The transcriptome start site of each gene was padded by 100bp to increase alignment rates  since reads were generated from the 5’ end of the transcript  and not all transcripts had correct annotations for their 5’ end. Additionally  pseudogenes were removed from the transcriptome as they sometimes interfered with real genes  given our 5’ coordinate selection described below. We chose the best alignment per read as the read with the most 5’ alignment coordinate to a transcript among those that aligned to the correct strand and had an alignment score of no more than 4 below the best alignment score overall. We then iteratively collapsed pairs of UMIs  if the sum of the minimum Phred quality scores across non matching bases was at most 30 to account for mis matched bases that were likely due to sequencing errors. A collapsed pair was represented for future comparisons by the UMI with a larger prior number of aligned reads unless the other UMI had at least 75% as many reads aligned and had equal or better UMI quality scores for those alignments evaluated as the softmin over aligned reads of the softmax over quality scores for the UMI bases in an aligned read. post collapsing  the number of transcripts observed per gene was calculated. This value was then adjusted based on the probability of a collision the drawing at random of the same UMI multiple times  which was calculated from the number of UMIs not detected for a given gene. The full length fragmented reads without xxx’ enrichment or UMI were aligned using RSEM against a reference transcriptome  which is depleted of mtRNA  rRNA  and pseudogenes as described above. The command ‘rsem calculate expression’ was used with parameters  p 4   paired end   seed length 24 for performing the alignment. The resulting gene levels in FPKM was used to build the expression matrix. Genome build: GRCz10 dr81 Supplementary files format and content: Tab delimited text files include gene levels UMI counts or FKPM for each sample.,Embryonic cell,,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for 5’ enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,Fish of each genotype were incrossed and embryos were collected. Embryos were dechorionated and allowed to grow at 28°C in standard fish water containing methylene blue till 50% epiboly stage.,developmental stage:50% epiboly|library preparation batch:p3|genotype:Wild type TLAB,GSM2838435,GSM2838435: wt p3 S19; Danio rerio; RNA Seq,GSM2838435,,1,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for five prime enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina HiSeq 2500,,SRP123526,,,wt_p3_S19.R1.fastq.gz wt_p3_S19.R2.fastq.gz,fastq fastq,60215008.0,1075268.0,GSM2838435 r1,0:24 1:32,A:16139361;C:13746582;G:14054555;T:16224611;N:49899,24,32,,,16139361,13746582,14054555,16224611,49899,SRX3357274,SRS2656356,SRA627905,GEO,"Schier Lab, Molecular and Cellular Biology, Harvard University",2,0.77939,0.83244,0.11193,0.11909,0.88667,0.88824,0.72767,0.71662,24,32,B,B,biological fallback assumption,illumina,hiseq_era,full_length,poly_a,unknown,sc,single_cell_plate,smartseq,,United States,2017-11-02,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
44102,SRR6250348,SRX3357273,SRS2656355,SRP123526,PRJNA416939,Single cell RNAseq SMART seq2 of wild type TLAB and MZoep tz57 zebrafish embryos at 50% epiboly stage,GSE106466,Transcriptome Analysis,SMART seq2 was performed on single cells isolated from visually staged zebrafish embryos. Overall design: Samples were all sequenced in one batch. Some were generated with a five prime' UMI tagged method  and others are full length SMART seq2.,parent bioproject:PRJNA417291,pubmed:29700225,,wt p3 S18,GSM2838434,,tissue:Embryonic cell|developmental stage:50% epiboly|library preparation batch:p3|genotype:Wild type TLAB,wt p3 S18,Bases 1 8 the UMI and template switch were trimmed from the remaining reads  trimmed stretches of polyA or polyT longer than 6bp and removed any reads that were shorter than 15 bp post trimming. We identified and depleted mtRNA and rRNA reads from the remaining reads by using Bowtie2 to align all reads against a transcriptome of mtRNA and rRNA  using the parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed   no unal. Remaining reads were then aligned to a modified version of the Ensembl Zv10 release 81 zebrafish transcriptome using Bowtie2 with parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed. The transcriptome start site of each gene was padded by 100bp to increase alignment rates  since reads were generated from the 5’ end of the transcript  and not all transcripts had correct annotations for their 5’ end. Additionally  pseudogenes were removed from the transcriptome as they sometimes interfered with real genes  given our 5’ coordinate selection described below. We chose the best alignment per read as the read with the most 5’ alignment coordinate to a transcript among those that aligned to the correct strand and had an alignment score of no more than 4 below the best alignment score overall. We then iteratively collapsed pairs of UMIs  if the sum of the minimum Phred quality scores across non matching bases was at most 30 to account for mis matched bases that were likely due to sequencing errors. A collapsed pair was represented for future comparisons by the UMI with a larger prior number of aligned reads unless the other UMI had at least 75% as many reads aligned and had equal or better UMI quality scores for those alignments evaluated as the softmin over aligned reads of the softmax over quality scores for the UMI bases in an aligned read. post collapsing  the number of transcripts observed per gene was calculated. This value was then adjusted based on the probability of a collision the drawing at random of the same UMI multiple times  which was calculated from the number of UMIs not detected for a given gene. The full length fragmented reads without xxx’ enrichment or UMI were aligned using RSEM against a reference transcriptome  which is depleted of mtRNA  rRNA  and pseudogenes as described above. The command ‘rsem calculate expression’ was used with parameters  p 4   paired end   seed length 24 for performing the alignment. The resulting gene levels in FPKM was used to build the expression matrix. Genome build: GRCz10 dr81 Supplementary files format and content: Tab delimited text files include gene levels UMI counts or FKPM for each sample.,Embryonic cell,,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for 5’ enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,Fish of each genotype were incrossed and embryos were collected. Embryos were dechorionated and allowed to grow at 28°C in standard fish water containing methylene blue till 50% epiboly stage.,developmental stage:50% epiboly|library preparation batch:p3|genotype:Wild type TLAB,GSM2838434,GSM2838434: wt p3 S18; Danio rerio; RNA Seq,GSM2838434,,1,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for five prime enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina HiSeq 2500,,SRP123526,,,wt_p3_S18.R1.fastq.gz wt_p3_S18.R2.fastq.gz,fastq fastq,86502416.0,1544686.0,GSM2838434 r1,0:24 1:32,A:22690851;C:20247122;G:20556837;T:22936501;N:71105,24,32,,,22690851,20247122,20556837,22936501,71105,SRX3357273,SRS2656355,SRA627905,GEO,"Schier Lab, Molecular and Cellular Biology, Harvard University",2,0.68812,0.73173,0.16183,0.17229,0.86748,0.8674,0.70363,0.69472,24,32,B,B,biological fallback assumption,illumina,hiseq_era,full_length,poly_a,unknown,sc,single_cell_plate,smartseq,,United States,2017-11-02,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
44103,SRR6250347,SRX3357271,SRS2656353,SRP123526,PRJNA416939,Single cell RNAseq SMART seq2 of wild type TLAB and MZoep tz57 zebrafish embryos at 50% epiboly stage,GSE106466,Transcriptome Analysis,SMART seq2 was performed on single cells isolated from visually staged zebrafish embryos. Overall design: Samples were all sequenced in one batch. Some were generated with a five prime' UMI tagged method  and others are full length SMART seq2.,parent bioproject:PRJNA417291,pubmed:29700225,,wt p3 S17,GSM2838433,,tissue:Embryonic cell|developmental stage:50% epiboly|library preparation batch:p3|genotype:Wild type TLAB,wt p3 S17,Bases 1 8 the UMI and template switch were trimmed from the remaining reads  trimmed stretches of polyA or polyT longer than 6bp and removed any reads that were shorter than 15 bp post trimming. We identified and depleted mtRNA and rRNA reads from the remaining reads by using Bowtie2 to align all reads against a transcriptome of mtRNA and rRNA  using the parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed   no unal. Remaining reads were then aligned to a modified version of the Ensembl Zv10 release 81 zebrafish transcriptome using Bowtie2 with parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed. The transcriptome start site of each gene was padded by 100bp to increase alignment rates  since reads were generated from the 5’ end of the transcript  and not all transcripts had correct annotations for their 5’ end. Additionally  pseudogenes were removed from the transcriptome as they sometimes interfered with real genes  given our 5’ coordinate selection described below. We chose the best alignment per read as the read with the most 5’ alignment coordinate to a transcript among those that aligned to the correct strand and had an alignment score of no more than 4 below the best alignment score overall. We then iteratively collapsed pairs of UMIs  if the sum of the minimum Phred quality scores across non matching bases was at most 30 to account for mis matched bases that were likely due to sequencing errors. A collapsed pair was represented for future comparisons by the UMI with a larger prior number of aligned reads unless the other UMI had at least 75% as many reads aligned and had equal or better UMI quality scores for those alignments evaluated as the softmin over aligned reads of the softmax over quality scores for the UMI bases in an aligned read. post collapsing  the number of transcripts observed per gene was calculated. This value was then adjusted based on the probability of a collision the drawing at random of the same UMI multiple times  which was calculated from the number of UMIs not detected for a given gene. The full length fragmented reads without xxx’ enrichment or UMI were aligned using RSEM against a reference transcriptome  which is depleted of mtRNA  rRNA  and pseudogenes as described above. The command ‘rsem calculate expression’ was used with parameters  p 4   paired end   seed length 24 for performing the alignment. The resulting gene levels in FPKM was used to build the expression matrix. Genome build: GRCz10 dr81 Supplementary files format and content: Tab delimited text files include gene levels UMI counts or FKPM for each sample.,Embryonic cell,,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for 5’ enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,Fish of each genotype were incrossed and embryos were collected. Embryos were dechorionated and allowed to grow at 28°C in standard fish water containing methylene blue till 50% epiboly stage.,developmental stage:50% epiboly|library preparation batch:p3|genotype:Wild type TLAB,GSM2838433,GSM2838433: wt p3 S17; Danio rerio; RNA Seq,GSM2838433,,1,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for five prime enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina HiSeq 2500,,SRP123526,,,wt_p3_S17.R1.fastq.gz wt_p3_S17.R2.fastq.gz,fastq fastq,85683024.0,1530054.0,GSM2838433 r1,0:24 1:32,A:22681508;C:19929532;G:20246491;T:22753269;N:72224,24,32,,,22681508,19929532,20246491,22753269,72224,SRX3357271,SRS2656353,SRA627905,GEO,"Schier Lab, Molecular and Cellular Biology, Harvard University",2,0.78043,0.82981,0.16296,0.17534,0.86856,0.86851,0.73198,0.73537,24,32,B,B,biological fallback assumption,illumina,hiseq_era,full_length,poly_a,unknown,sc,single_cell_plate,smartseq,,United States,2017-11-02,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
44104,SRR6250346,SRX3357270,SRS2656354,SRP123526,PRJNA416939,Single cell RNAseq SMART seq2 of wild type TLAB and MZoep tz57 zebrafish embryos at 50% epiboly stage,GSE106466,Transcriptome Analysis,SMART seq2 was performed on single cells isolated from visually staged zebrafish embryos. Overall design: Samples were all sequenced in one batch. Some were generated with a five prime' UMI tagged method  and others are full length SMART seq2.,parent bioproject:PRJNA417291,pubmed:29700225,,wt p3 S16,GSM2838432,,tissue:Embryonic cell|developmental stage:50% epiboly|library preparation batch:p3|genotype:Wild type TLAB,wt p3 S16,Bases 1 8 the UMI and template switch were trimmed from the remaining reads  trimmed stretches of polyA or polyT longer than 6bp and removed any reads that were shorter than 15 bp post trimming. We identified and depleted mtRNA and rRNA reads from the remaining reads by using Bowtie2 to align all reads against a transcriptome of mtRNA and rRNA  using the parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed   no unal. Remaining reads were then aligned to a modified version of the Ensembl Zv10 release 81 zebrafish transcriptome using Bowtie2 with parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed. The transcriptome start site of each gene was padded by 100bp to increase alignment rates  since reads were generated from the 5’ end of the transcript  and not all transcripts had correct annotations for their 5’ end. Additionally  pseudogenes were removed from the transcriptome as they sometimes interfered with real genes  given our 5’ coordinate selection described below. We chose the best alignment per read as the read with the most 5’ alignment coordinate to a transcript among those that aligned to the correct strand and had an alignment score of no more than 4 below the best alignment score overall. We then iteratively collapsed pairs of UMIs  if the sum of the minimum Phred quality scores across non matching bases was at most 30 to account for mis matched bases that were likely due to sequencing errors. A collapsed pair was represented for future comparisons by the UMI with a larger prior number of aligned reads unless the other UMI had at least 75% as many reads aligned and had equal or better UMI quality scores for those alignments evaluated as the softmin over aligned reads of the softmax over quality scores for the UMI bases in an aligned read. post collapsing  the number of transcripts observed per gene was calculated. This value was then adjusted based on the probability of a collision the drawing at random of the same UMI multiple times  which was calculated from the number of UMIs not detected for a given gene. The full length fragmented reads without xxx’ enrichment or UMI were aligned using RSEM against a reference transcriptome  which is depleted of mtRNA  rRNA  and pseudogenes as described above. The command ‘rsem calculate expression’ was used with parameters  p 4   paired end   seed length 24 for performing the alignment. The resulting gene levels in FPKM was used to build the expression matrix. Genome build: GRCz10 dr81 Supplementary files format and content: Tab delimited text files include gene levels UMI counts or FKPM for each sample.,Embryonic cell,,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for 5’ enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,Fish of each genotype were incrossed and embryos were collected. Embryos were dechorionated and allowed to grow at 28°C in standard fish water containing methylene blue till 50% epiboly stage.,developmental stage:50% epiboly|library preparation batch:p3|genotype:Wild type TLAB,GSM2838432,GSM2838432: wt p3 S16; Danio rerio; RNA Seq,GSM2838432,,1,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for five prime enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina HiSeq 2500,,SRP123526,,,wt_p3_S16.R1.fastq.gz wt_p3_S16.R2.fastq.gz,fastq fastq,76375544.0,1363849.0,GSM2838432 r1,0:24 1:32,A:19872118;C:18067491;G:18341353;T:20032879;N:61703,24,32,,,19872118,18067491,18341353,20032879,61703,SRX3357270,SRS2656354,SRA627905,GEO,"Schier Lab, Molecular and Cellular Biology, Harvard University",2,0.78228,0.83295,0.18441,0.19896,0.86691,0.86797,0.7169,0.70775,24,32,B,B,biological fallback assumption,illumina,hiseq_era,full_length,poly_a,unknown,sc,single_cell_plate,smartseq,,United States,2017-11-02,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
44105,SRR6250345,SRX3357269,SRS2656351,SRP123526,PRJNA416939,Single cell RNAseq SMART seq2 of wild type TLAB and MZoep tz57 zebrafish embryos at 50% epiboly stage,GSE106466,Transcriptome Analysis,SMART seq2 was performed on single cells isolated from visually staged zebrafish embryos. Overall design: Samples were all sequenced in one batch. Some were generated with a five prime' UMI tagged method  and others are full length SMART seq2.,parent bioproject:PRJNA417291,pubmed:29700225,,wt p3 S15,GSM2838431,,tissue:Embryonic cell|developmental stage:50% epiboly|library preparation batch:p3|genotype:Wild type TLAB,wt p3 S15,Bases 1 8 the UMI and template switch were trimmed from the remaining reads  trimmed stretches of polyA or polyT longer than 6bp and removed any reads that were shorter than 15 bp post trimming. We identified and depleted mtRNA and rRNA reads from the remaining reads by using Bowtie2 to align all reads against a transcriptome of mtRNA and rRNA  using the parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed   no unal. Remaining reads were then aligned to a modified version of the Ensembl Zv10 release 81 zebrafish transcriptome using Bowtie2 with parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed. The transcriptome start site of each gene was padded by 100bp to increase alignment rates  since reads were generated from the 5’ end of the transcript  and not all transcripts had correct annotations for their 5’ end. Additionally  pseudogenes were removed from the transcriptome as they sometimes interfered with real genes  given our 5’ coordinate selection described below. We chose the best alignment per read as the read with the most 5’ alignment coordinate to a transcript among those that aligned to the correct strand and had an alignment score of no more than 4 below the best alignment score overall. We then iteratively collapsed pairs of UMIs  if the sum of the minimum Phred quality scores across non matching bases was at most 30 to account for mis matched bases that were likely due to sequencing errors. A collapsed pair was represented for future comparisons by the UMI with a larger prior number of aligned reads unless the other UMI had at least 75% as many reads aligned and had equal or better UMI quality scores for those alignments evaluated as the softmin over aligned reads of the softmax over quality scores for the UMI bases in an aligned read. post collapsing  the number of transcripts observed per gene was calculated. This value was then adjusted based on the probability of a collision the drawing at random of the same UMI multiple times  which was calculated from the number of UMIs not detected for a given gene. The full length fragmented reads without xxx’ enrichment or UMI were aligned using RSEM against a reference transcriptome  which is depleted of mtRNA  rRNA  and pseudogenes as described above. The command ‘rsem calculate expression’ was used with parameters  p 4   paired end   seed length 24 for performing the alignment. The resulting gene levels in FPKM was used to build the expression matrix. Genome build: GRCz10 dr81 Supplementary files format and content: Tab delimited text files include gene levels UMI counts or FKPM for each sample.,Embryonic cell,,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for 5’ enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,Fish of each genotype were incrossed and embryos were collected. Embryos were dechorionated and allowed to grow at 28°C in standard fish water containing methylene blue till 50% epiboly stage.,developmental stage:50% epiboly|library preparation batch:p3|genotype:Wild type TLAB,GSM2838431,GSM2838431: wt p3 S15; Danio rerio; RNA Seq,GSM2838431,,1,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for five prime enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina HiSeq 2500,,SRP123526,,,wt_p3_S15.R1.fastq.gz wt_p3_S15.R2.fastq.gz,fastq fastq,77270816.0,1379836.0,GSM2838431 r1,0:24 1:32,A:20604868;C:17888374;G:18173946;T:20541155;N:62473,24,32,,,20604868,17888374,18173946,20541155,62473,SRX3357269,SRS2656351,SRA627905,GEO,"Schier Lab, Molecular and Cellular Biology, Harvard University",2,0.78034,0.83192,0.15119,0.16258,0.86943,0.87004,0.73186,0.73196,24,32,B,B,biological fallback assumption,illumina,hiseq_era,full_length,poly_a,unknown,sc,single_cell_plate,smartseq,,United States,2017-11-02,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
44106,SRR6250344,SRX3357268,SRS2656350,SRP123526,PRJNA416939,Single cell RNAseq SMART seq2 of wild type TLAB and MZoep tz57 zebrafish embryos at 50% epiboly stage,GSE106466,Transcriptome Analysis,SMART seq2 was performed on single cells isolated from visually staged zebrafish embryos. Overall design: Samples were all sequenced in one batch. Some were generated with a five prime' UMI tagged method  and others are full length SMART seq2.,parent bioproject:PRJNA417291,pubmed:29700225,,wt p3 S14,GSM2838430,,tissue:Embryonic cell|developmental stage:50% epiboly|library preparation batch:p3|genotype:Wild type TLAB,wt p3 S14,Bases 1 8 the UMI and template switch were trimmed from the remaining reads  trimmed stretches of polyA or polyT longer than 6bp and removed any reads that were shorter than 15 bp post trimming. We identified and depleted mtRNA and rRNA reads from the remaining reads by using Bowtie2 to align all reads against a transcriptome of mtRNA and rRNA  using the parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed   no unal. Remaining reads were then aligned to a modified version of the Ensembl Zv10 release 81 zebrafish transcriptome using Bowtie2 with parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed. The transcriptome start site of each gene was padded by 100bp to increase alignment rates  since reads were generated from the 5’ end of the transcript  and not all transcripts had correct annotations for their 5’ end. Additionally  pseudogenes were removed from the transcriptome as they sometimes interfered with real genes  given our 5’ coordinate selection described below. We chose the best alignment per read as the read with the most 5’ alignment coordinate to a transcript among those that aligned to the correct strand and had an alignment score of no more than 4 below the best alignment score overall. We then iteratively collapsed pairs of UMIs  if the sum of the minimum Phred quality scores across non matching bases was at most 30 to account for mis matched bases that were likely due to sequencing errors. A collapsed pair was represented for future comparisons by the UMI with a larger prior number of aligned reads unless the other UMI had at least 75% as many reads aligned and had equal or better UMI quality scores for those alignments evaluated as the softmin over aligned reads of the softmax over quality scores for the UMI bases in an aligned read. post collapsing  the number of transcripts observed per gene was calculated. This value was then adjusted based on the probability of a collision the drawing at random of the same UMI multiple times  which was calculated from the number of UMIs not detected for a given gene. The full length fragmented reads without xxx’ enrichment or UMI were aligned using RSEM against a reference transcriptome  which is depleted of mtRNA  rRNA  and pseudogenes as described above. The command ‘rsem calculate expression’ was used with parameters  p 4   paired end   seed length 24 for performing the alignment. The resulting gene levels in FPKM was used to build the expression matrix. Genome build: GRCz10 dr81 Supplementary files format and content: Tab delimited text files include gene levels UMI counts or FKPM for each sample.,Embryonic cell,,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for 5’ enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,Fish of each genotype were incrossed and embryos were collected. Embryos were dechorionated and allowed to grow at 28°C in standard fish water containing methylene blue till 50% epiboly stage.,developmental stage:50% epiboly|library preparation batch:p3|genotype:Wild type TLAB,GSM2838430,GSM2838430: wt p3 S14; Danio rerio; RNA Seq,GSM2838430,,1,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for five prime enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina HiSeq 2500,,SRP123526,,,wt_p3_S14.R1.fastq.gz wt_p3_S14.R2.fastq.gz,fastq fastq,72452464.0,1293794.0,GSM2838430 r1,0:24 1:32,A:19506576;C:16600430;G:16874190;T:19412739;N:58529,24,32,,,19506576,16600430,16874190,19412739,58529,SRX3357268,SRS2656350,SRA627905,GEO,"Schier Lab, Molecular and Cellular Biology, Harvard University",2,0.77447,0.82555,0.13844,0.1491,0.88915,0.88939,0.77862,0.78435,24,32,B,B,biological fallback assumption,illumina,hiseq_era,full_length,poly_a,unknown,sc,single_cell_plate,smartseq,,United States,2017-11-02,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
44107,SRR6250343,SRX3357267,SRS2656349,SRP123526,PRJNA416939,Single cell RNAseq SMART seq2 of wild type TLAB and MZoep tz57 zebrafish embryos at 50% epiboly stage,GSE106466,Transcriptome Analysis,SMART seq2 was performed on single cells isolated from visually staged zebrafish embryos. Overall design: Samples were all sequenced in one batch. Some were generated with a five prime' UMI tagged method  and others are full length SMART seq2.,parent bioproject:PRJNA417291,pubmed:29700225,,wt p3 S13,GSM2838429,,tissue:Embryonic cell|developmental stage:50% epiboly|library preparation batch:p3|genotype:Wild type TLAB,wt p3 S13,Bases 1 8 the UMI and template switch were trimmed from the remaining reads  trimmed stretches of polyA or polyT longer than 6bp and removed any reads that were shorter than 15 bp post trimming. We identified and depleted mtRNA and rRNA reads from the remaining reads by using Bowtie2 to align all reads against a transcriptome of mtRNA and rRNA  using the parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed   no unal. Remaining reads were then aligned to a modified version of the Ensembl Zv10 release 81 zebrafish transcriptome using Bowtie2 with parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed. The transcriptome start site of each gene was padded by 100bp to increase alignment rates  since reads were generated from the 5’ end of the transcript  and not all transcripts had correct annotations for their 5’ end. Additionally  pseudogenes were removed from the transcriptome as they sometimes interfered with real genes  given our 5’ coordinate selection described below. We chose the best alignment per read as the read with the most 5’ alignment coordinate to a transcript among those that aligned to the correct strand and had an alignment score of no more than 4 below the best alignment score overall. We then iteratively collapsed pairs of UMIs  if the sum of the minimum Phred quality scores across non matching bases was at most 30 to account for mis matched bases that were likely due to sequencing errors. A collapsed pair was represented for future comparisons by the UMI with a larger prior number of aligned reads unless the other UMI had at least 75% as many reads aligned and had equal or better UMI quality scores for those alignments evaluated as the softmin over aligned reads of the softmax over quality scores for the UMI bases in an aligned read. post collapsing  the number of transcripts observed per gene was calculated. This value was then adjusted based on the probability of a collision the drawing at random of the same UMI multiple times  which was calculated from the number of UMIs not detected for a given gene. The full length fragmented reads without xxx’ enrichment or UMI were aligned using RSEM against a reference transcriptome  which is depleted of mtRNA  rRNA  and pseudogenes as described above. The command ‘rsem calculate expression’ was used with parameters  p 4   paired end   seed length 24 for performing the alignment. The resulting gene levels in FPKM was used to build the expression matrix. Genome build: GRCz10 dr81 Supplementary files format and content: Tab delimited text files include gene levels UMI counts or FKPM for each sample.,Embryonic cell,,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for 5’ enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,Fish of each genotype were incrossed and embryos were collected. Embryos were dechorionated and allowed to grow at 28°C in standard fish water containing methylene blue till 50% epiboly stage.,developmental stage:50% epiboly|library preparation batch:p3|genotype:Wild type TLAB,GSM2838429,GSM2838429: wt p3 S13; Danio rerio; RNA Seq,GSM2838429,,1,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for five prime enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina HiSeq 2500,,SRP123526,,,wt_p3_S13.R1.fastq.gz wt_p3_S13.R2.fastq.gz,fastq fastq,77600880.0,1385730.0,GSM2838429 r1,0:24 1:32,A:20408069;C:18159300;G:18461922;T:20508697;N:62892,24,32,,,20408069,18159300,18461922,20508697,62892,SRX3357267,SRS2656349,SRA627905,GEO,"Schier Lab, Molecular and Cellular Biology, Harvard University",2,0.7812,0.83306,0.18183,0.19603,0.86225,0.86373,0.70873,0.70613,24,32,B,B,biological fallback assumption,illumina,hiseq_era,full_length,poly_a,unknown,sc,single_cell_plate,smartseq,,United States,2017-11-02,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
44108,SRR6250342,SRX3357266,SRS2656348,SRP123526,PRJNA416939,Single cell RNAseq SMART seq2 of wild type TLAB and MZoep tz57 zebrafish embryos at 50% epiboly stage,GSE106466,Transcriptome Analysis,SMART seq2 was performed on single cells isolated from visually staged zebrafish embryos. Overall design: Samples were all sequenced in one batch. Some were generated with a five prime' UMI tagged method  and others are full length SMART seq2.,parent bioproject:PRJNA417291,pubmed:29700225,,wt p3 S12,GSM2838428,,tissue:Embryonic cell|developmental stage:50% epiboly|library preparation batch:p3|genotype:Wild type TLAB,wt p3 S12,Bases 1 8 the UMI and template switch were trimmed from the remaining reads  trimmed stretches of polyA or polyT longer than 6bp and removed any reads that were shorter than 15 bp post trimming. We identified and depleted mtRNA and rRNA reads from the remaining reads by using Bowtie2 to align all reads against a transcriptome of mtRNA and rRNA  using the parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed   no unal. Remaining reads were then aligned to a modified version of the Ensembl Zv10 release 81 zebrafish transcriptome using Bowtie2 with parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed. The transcriptome start site of each gene was padded by 100bp to increase alignment rates  since reads were generated from the 5’ end of the transcript  and not all transcripts had correct annotations for their 5’ end. Additionally  pseudogenes were removed from the transcriptome as they sometimes interfered with real genes  given our 5’ coordinate selection described below. We chose the best alignment per read as the read with the most 5’ alignment coordinate to a transcript among those that aligned to the correct strand and had an alignment score of no more than 4 below the best alignment score overall. We then iteratively collapsed pairs of UMIs  if the sum of the minimum Phred quality scores across non matching bases was at most 30 to account for mis matched bases that were likely due to sequencing errors. A collapsed pair was represented for future comparisons by the UMI with a larger prior number of aligned reads unless the other UMI had at least 75% as many reads aligned and had equal or better UMI quality scores for those alignments evaluated as the softmin over aligned reads of the softmax over quality scores for the UMI bases in an aligned read. post collapsing  the number of transcripts observed per gene was calculated. This value was then adjusted based on the probability of a collision the drawing at random of the same UMI multiple times  which was calculated from the number of UMIs not detected for a given gene. The full length fragmented reads without xxx’ enrichment or UMI were aligned using RSEM against a reference transcriptome  which is depleted of mtRNA  rRNA  and pseudogenes as described above. The command ‘rsem calculate expression’ was used with parameters  p 4   paired end   seed length 24 for performing the alignment. The resulting gene levels in FPKM was used to build the expression matrix. Genome build: GRCz10 dr81 Supplementary files format and content: Tab delimited text files include gene levels UMI counts or FKPM for each sample.,Embryonic cell,,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for 5’ enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,Fish of each genotype were incrossed and embryos were collected. Embryos were dechorionated and allowed to grow at 28°C in standard fish water containing methylene blue till 50% epiboly stage.,developmental stage:50% epiboly|library preparation batch:p3|genotype:Wild type TLAB,GSM2838428,GSM2838428: wt p3 S12; Danio rerio; RNA Seq,GSM2838428,,1,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for five prime enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina HiSeq 2500,,SRP123526,,,wt_p3_S12.R1.fastq.gz wt_p3_S12.R2.fastq.gz,fastq fastq,67502568.0,1205403.0,GSM2838428 r1,0:24 1:32,A:18195431;C:15418993;G:15695230;T:18137911;N:55003,24,32,,,18195431,15418993,15695230,18137911,55003,SRX3357266,SRS2656348,SRA627905,GEO,"Schier Lab, Molecular and Cellular Biology, Harvard University",2,0.77379,0.82796,0.13406,0.14349,0.88256,0.88258,0.76096,0.7618,24,32,B,B,biological fallback assumption,illumina,hiseq_era,full_length,poly_a,unknown,sc,single_cell_plate,smartseq,,United States,2017-11-02,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
44109,SRR6250341,SRX3357265,SRS2656347,SRP123526,PRJNA416939,Single cell RNAseq SMART seq2 of wild type TLAB and MZoep tz57 zebrafish embryos at 50% epiboly stage,GSE106466,Transcriptome Analysis,SMART seq2 was performed on single cells isolated from visually staged zebrafish embryos. Overall design: Samples were all sequenced in one batch. Some were generated with a five prime' UMI tagged method  and others are full length SMART seq2.,parent bioproject:PRJNA417291,pubmed:29700225,,wt p3 S11,GSM2838427,,tissue:Embryonic cell|developmental stage:50% epiboly|library preparation batch:p3|genotype:Wild type TLAB,wt p3 S11,Bases 1 8 the UMI and template switch were trimmed from the remaining reads  trimmed stretches of polyA or polyT longer than 6bp and removed any reads that were shorter than 15 bp post trimming. We identified and depleted mtRNA and rRNA reads from the remaining reads by using Bowtie2 to align all reads against a transcriptome of mtRNA and rRNA  using the parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed   no unal. Remaining reads were then aligned to a modified version of the Ensembl Zv10 release 81 zebrafish transcriptome using Bowtie2 with parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed. The transcriptome start site of each gene was padded by 100bp to increase alignment rates  since reads were generated from the 5’ end of the transcript  and not all transcripts had correct annotations for their 5’ end. Additionally  pseudogenes were removed from the transcriptome as they sometimes interfered with real genes  given our 5’ coordinate selection described below. We chose the best alignment per read as the read with the most 5’ alignment coordinate to a transcript among those that aligned to the correct strand and had an alignment score of no more than 4 below the best alignment score overall. We then iteratively collapsed pairs of UMIs  if the sum of the minimum Phred quality scores across non matching bases was at most 30 to account for mis matched bases that were likely due to sequencing errors. A collapsed pair was represented for future comparisons by the UMI with a larger prior number of aligned reads unless the other UMI had at least 75% as many reads aligned and had equal or better UMI quality scores for those alignments evaluated as the softmin over aligned reads of the softmax over quality scores for the UMI bases in an aligned read. post collapsing  the number of transcripts observed per gene was calculated. This value was then adjusted based on the probability of a collision the drawing at random of the same UMI multiple times  which was calculated from the number of UMIs not detected for a given gene. The full length fragmented reads without xxx’ enrichment or UMI were aligned using RSEM against a reference transcriptome  which is depleted of mtRNA  rRNA  and pseudogenes as described above. The command ‘rsem calculate expression’ was used with parameters  p 4   paired end   seed length 24 for performing the alignment. The resulting gene levels in FPKM was used to build the expression matrix. Genome build: GRCz10 dr81 Supplementary files format and content: Tab delimited text files include gene levels UMI counts or FKPM for each sample.,Embryonic cell,,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for 5’ enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,Fish of each genotype were incrossed and embryos were collected. Embryos were dechorionated and allowed to grow at 28°C in standard fish water containing methylene blue till 50% epiboly stage.,developmental stage:50% epiboly|library preparation batch:p3|genotype:Wild type TLAB,GSM2838427,GSM2838427: wt p3 S11; Danio rerio; RNA Seq,GSM2838427,,1,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for five prime enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina HiSeq 2500,,SRP123526,,,wt_p3_S11.R1.fastq.gz wt_p3_S11.R2.fastq.gz,fastq fastq,2408.0,43.0,GSM2838427 r1,0:24 1:32,A:568;C:630;G:651;T:559;N:0,24,32,,,568,630,651,559,0,SRX3357265,SRS2656347,SRA627905,GEO,"Schier Lab, Molecular and Cellular Biology, Harvard University",2,0.7742,0.38462,0.35483,0.07692,0.99977,0.99991,0.6923,1.0,24,32,B,B,mate2-mate1 similar by mapping diff,illumina,hiseq_era,full_length,poly_a,unknown,sc,single_cell_plate,smartseq,,United States,2017-11-02,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
44110,SRR6250340,SRX3357264,SRS2656346,SRP123526,PRJNA416939,Single cell RNAseq SMART seq2 of wild type TLAB and MZoep tz57 zebrafish embryos at 50% epiboly stage,GSE106466,Transcriptome Analysis,SMART seq2 was performed on single cells isolated from visually staged zebrafish embryos. Overall design: Samples were all sequenced in one batch. Some were generated with a five prime' UMI tagged method  and others are full length SMART seq2.,parent bioproject:PRJNA417291,pubmed:29700225,,wt p3 S10,GSM2838426,,tissue:Embryonic cell|developmental stage:50% epiboly|library preparation batch:p3|genotype:Wild type TLAB,wt p3 S10,Bases 1 8 the UMI and template switch were trimmed from the remaining reads  trimmed stretches of polyA or polyT longer than 6bp and removed any reads that were shorter than 15 bp post trimming. We identified and depleted mtRNA and rRNA reads from the remaining reads by using Bowtie2 to align all reads against a transcriptome of mtRNA and rRNA  using the parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed   no unal. Remaining reads were then aligned to a modified version of the Ensembl Zv10 release 81 zebrafish transcriptome using Bowtie2 with parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed. The transcriptome start site of each gene was padded by 100bp to increase alignment rates  since reads were generated from the 5’ end of the transcript  and not all transcripts had correct annotations for their 5’ end. Additionally  pseudogenes were removed from the transcriptome as they sometimes interfered with real genes  given our 5’ coordinate selection described below. We chose the best alignment per read as the read with the most 5’ alignment coordinate to a transcript among those that aligned to the correct strand and had an alignment score of no more than 4 below the best alignment score overall. We then iteratively collapsed pairs of UMIs  if the sum of the minimum Phred quality scores across non matching bases was at most 30 to account for mis matched bases that were likely due to sequencing errors. A collapsed pair was represented for future comparisons by the UMI with a larger prior number of aligned reads unless the other UMI had at least 75% as many reads aligned and had equal or better UMI quality scores for those alignments evaluated as the softmin over aligned reads of the softmax over quality scores for the UMI bases in an aligned read. post collapsing  the number of transcripts observed per gene was calculated. This value was then adjusted based on the probability of a collision the drawing at random of the same UMI multiple times  which was calculated from the number of UMIs not detected for a given gene. The full length fragmented reads without xxx’ enrichment or UMI were aligned using RSEM against a reference transcriptome  which is depleted of mtRNA  rRNA  and pseudogenes as described above. The command ‘rsem calculate expression’ was used with parameters  p 4   paired end   seed length 24 for performing the alignment. The resulting gene levels in FPKM was used to build the expression matrix. Genome build: GRCz10 dr81 Supplementary files format and content: Tab delimited text files include gene levels UMI counts or FKPM for each sample.,Embryonic cell,,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for 5’ enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,Fish of each genotype were incrossed and embryos were collected. Embryos were dechorionated and allowed to grow at 28°C in standard fish water containing methylene blue till 50% epiboly stage.,developmental stage:50% epiboly|library preparation batch:p3|genotype:Wild type TLAB,GSM2838426,GSM2838426: wt p3 S10; Danio rerio; RNA Seq,GSM2838426,,1,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for five prime enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina HiSeq 2500,,SRP123526,,,wt_p3_S10.R1.fastq.gz wt_p3_S10.R2.fastq.gz,fastq fastq,1064.0,19.0,GSM2838426 r1,0:24 1:32,A:242;C:270;G:291;T:261;N:0,24,32,,,242,270,291,261,0,SRX3357264,SRS2656346,SRA627905,GEO,"Schier Lab, Molecular and Cellular Biology, Harvard University",2,0.63637,0.66667,0.0,0.33333,0.99989,0.99997,0.71428,1.0,24,32,B,B,biological fallback assumption,illumina,hiseq_era,full_length,poly_a,unknown,sc,single_cell_plate,smartseq,,United States,2017-11-02,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
44111,SRR6250339,SRX3357263,SRS2656344,SRP123526,PRJNA416939,Single cell RNAseq SMART seq2 of wild type TLAB and MZoep tz57 zebrafish embryos at 50% epiboly stage,GSE106466,Transcriptome Analysis,SMART seq2 was performed on single cells isolated from visually staged zebrafish embryos. Overall design: Samples were all sequenced in one batch. Some were generated with a five prime' UMI tagged method  and others are full length SMART seq2.,parent bioproject:PRJNA417291,pubmed:29700225,,wt p2 S9,GSM2838425,,tissue:Embryonic cell|developmental stage:50% epiboly|library preparation batch:p2|genotype:Wild type TLAB,wt p2 S9,Bases 1 8 the UMI and template switch were trimmed from the remaining reads  trimmed stretches of polyA or polyT longer than 6bp and removed any reads that were shorter than 15 bp post trimming. We identified and depleted mtRNA and rRNA reads from the remaining reads by using Bowtie2 to align all reads against a transcriptome of mtRNA and rRNA  using the parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed   no unal. Remaining reads were then aligned to a modified version of the Ensembl Zv10 release 81 zebrafish transcriptome using Bowtie2 with parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed. The transcriptome start site of each gene was padded by 100bp to increase alignment rates  since reads were generated from the 5’ end of the transcript  and not all transcripts had correct annotations for their 5’ end. Additionally  pseudogenes were removed from the transcriptome as they sometimes interfered with real genes  given our 5’ coordinate selection described below. We chose the best alignment per read as the read with the most 5’ alignment coordinate to a transcript among those that aligned to the correct strand and had an alignment score of no more than 4 below the best alignment score overall. We then iteratively collapsed pairs of UMIs  if the sum of the minimum Phred quality scores across non matching bases was at most 30 to account for mis matched bases that were likely due to sequencing errors. A collapsed pair was represented for future comparisons by the UMI with a larger prior number of aligned reads unless the other UMI had at least 75% as many reads aligned and had equal or better UMI quality scores for those alignments evaluated as the softmin over aligned reads of the softmax over quality scores for the UMI bases in an aligned read. post collapsing  the number of transcripts observed per gene was calculated. This value was then adjusted based on the probability of a collision the drawing at random of the same UMI multiple times  which was calculated from the number of UMIs not detected for a given gene. The full length fragmented reads without xxx’ enrichment or UMI were aligned using RSEM against a reference transcriptome  which is depleted of mtRNA  rRNA  and pseudogenes as described above. The command ‘rsem calculate expression’ was used with parameters  p 4   paired end   seed length 24 for performing the alignment. The resulting gene levels in FPKM was used to build the expression matrix. Genome build: GRCz10 dr81 Supplementary files format and content: Tab delimited text files include gene levels UMI counts or FKPM for each sample.,Embryonic cell,,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for 5’ enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,Fish of each genotype were incrossed and embryos were collected. Embryos were dechorionated and allowed to grow at 28°C in standard fish water containing methylene blue till 50% epiboly stage.,developmental stage:50% epiboly|library preparation batch:p2|genotype:Wild type TLAB,GSM2838425,GSM2838425: wt p2 S9; Danio rerio; RNA Seq,GSM2838425,,1,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for five prime enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina HiSeq 2500,,SRP123526,,,wt_p2_S9.R1.fastq.gz wt_p2_S9.R2.fastq.gz,fastq fastq,40783792.0,728282.0,GSM2838425 r1,0:24 1:32,A:10649223;C:9569702;G:9637527;T:10894504;N:32836,24,32,,,10649223,9569702,9637527,10894504,32836,SRX3357263,SRS2656344,SRA627905,GEO,"Schier Lab, Molecular and Cellular Biology, Harvard University",2,0.7906,0.84558,0.20591,0.21966,0.85865,0.85898,0.71966,0.72194,24,32,B,B,biological fallback assumption,illumina,hiseq_era,full_length,poly_a,unknown,sc,single_cell_plate,smartseq,,United States,2017-11-02,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
44112,SRR6250338,SRX3357262,SRS2656343,SRP123526,PRJNA416939,Single cell RNAseq SMART seq2 of wild type TLAB and MZoep tz57 zebrafish embryos at 50% epiboly stage,GSE106466,Transcriptome Analysis,SMART seq2 was performed on single cells isolated from visually staged zebrafish embryos. Overall design: Samples were all sequenced in one batch. Some were generated with a five prime' UMI tagged method  and others are full length SMART seq2.,parent bioproject:PRJNA417291,pubmed:29700225,,wt p2 S8,GSM2838424,,tissue:Embryonic cell|developmental stage:50% epiboly|library preparation batch:p2|genotype:Wild type TLAB,wt p2 S8,Bases 1 8 the UMI and template switch were trimmed from the remaining reads  trimmed stretches of polyA or polyT longer than 6bp and removed any reads that were shorter than 15 bp post trimming. We identified and depleted mtRNA and rRNA reads from the remaining reads by using Bowtie2 to align all reads against a transcriptome of mtRNA and rRNA  using the parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed   no unal. Remaining reads were then aligned to a modified version of the Ensembl Zv10 release 81 zebrafish transcriptome using Bowtie2 with parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed. The transcriptome start site of each gene was padded by 100bp to increase alignment rates  since reads were generated from the 5’ end of the transcript  and not all transcripts had correct annotations for their 5’ end. Additionally  pseudogenes were removed from the transcriptome as they sometimes interfered with real genes  given our 5’ coordinate selection described below. We chose the best alignment per read as the read with the most 5’ alignment coordinate to a transcript among those that aligned to the correct strand and had an alignment score of no more than 4 below the best alignment score overall. We then iteratively collapsed pairs of UMIs  if the sum of the minimum Phred quality scores across non matching bases was at most 30 to account for mis matched bases that were likely due to sequencing errors. A collapsed pair was represented for future comparisons by the UMI with a larger prior number of aligned reads unless the other UMI had at least 75% as many reads aligned and had equal or better UMI quality scores for those alignments evaluated as the softmin over aligned reads of the softmax over quality scores for the UMI bases in an aligned read. post collapsing  the number of transcripts observed per gene was calculated. This value was then adjusted based on the probability of a collision the drawing at random of the same UMI multiple times  which was calculated from the number of UMIs not detected for a given gene. The full length fragmented reads without xxx’ enrichment or UMI were aligned using RSEM against a reference transcriptome  which is depleted of mtRNA  rRNA  and pseudogenes as described above. The command ‘rsem calculate expression’ was used with parameters  p 4   paired end   seed length 24 for performing the alignment. The resulting gene levels in FPKM was used to build the expression matrix. Genome build: GRCz10 dr81 Supplementary files format and content: Tab delimited text files include gene levels UMI counts or FKPM for each sample.,Embryonic cell,,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for 5’ enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,Fish of each genotype were incrossed and embryos were collected. Embryos were dechorionated and allowed to grow at 28°C in standard fish water containing methylene blue till 50% epiboly stage.,developmental stage:50% epiboly|library preparation batch:p2|genotype:Wild type TLAB,GSM2838424,GSM2838424: wt p2 S8; Danio rerio; RNA Seq,GSM2838424,,1,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for five prime enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina HiSeq 2500,,SRP123526,,,wt_p2_S8.R1.fastq.gz wt_p2_S8.R2.fastq.gz,fastq fastq,39089400.0,698025.0,GSM2838424 r1,,,,,,,,,,,,SRX3357262,SRS2656343,SRA627905,GEO,"Schier Lab, Molecular and Cellular Biology, Harvard University",2,0.78394,0.82511,0.20608,0.21878,0.86882,0.86998,0.69976,0.71133,24,32,B,B,biological fallback assumption,illumina,hiseq_era,full_length,poly_a,unknown,sc,single_cell_plate,smartseq,,United States,2017-11-02,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
44113,SRR6250337,SRX3357261,SRS2656342,SRP123526,PRJNA416939,Single cell RNAseq SMART seq2 of wild type TLAB and MZoep tz57 zebrafish embryos at 50% epiboly stage,GSE106466,Transcriptome Analysis,SMART seq2 was performed on single cells isolated from visually staged zebrafish embryos. Overall design: Samples were all sequenced in one batch. Some were generated with a five prime' UMI tagged method  and others are full length SMART seq2.,parent bioproject:PRJNA417291,pubmed:29700225,,wt p2 S7,GSM2838423,,tissue:Embryonic cell|developmental stage:50% epiboly|library preparation batch:p2|genotype:Wild type TLAB,wt p2 S7,Bases 1 8 the UMI and template switch were trimmed from the remaining reads  trimmed stretches of polyA or polyT longer than 6bp and removed any reads that were shorter than 15 bp post trimming. We identified and depleted mtRNA and rRNA reads from the remaining reads by using Bowtie2 to align all reads against a transcriptome of mtRNA and rRNA  using the parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed   no unal. Remaining reads were then aligned to a modified version of the Ensembl Zv10 release 81 zebrafish transcriptome using Bowtie2 with parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed. The transcriptome start site of each gene was padded by 100bp to increase alignment rates  since reads were generated from the 5’ end of the transcript  and not all transcripts had correct annotations for their 5’ end. Additionally  pseudogenes were removed from the transcriptome as they sometimes interfered with real genes  given our 5’ coordinate selection described below. We chose the best alignment per read as the read with the most 5’ alignment coordinate to a transcript among those that aligned to the correct strand and had an alignment score of no more than 4 below the best alignment score overall. We then iteratively collapsed pairs of UMIs  if the sum of the minimum Phred quality scores across non matching bases was at most 30 to account for mis matched bases that were likely due to sequencing errors. A collapsed pair was represented for future comparisons by the UMI with a larger prior number of aligned reads unless the other UMI had at least 75% as many reads aligned and had equal or better UMI quality scores for those alignments evaluated as the softmin over aligned reads of the softmax over quality scores for the UMI bases in an aligned read. post collapsing  the number of transcripts observed per gene was calculated. This value was then adjusted based on the probability of a collision the drawing at random of the same UMI multiple times  which was calculated from the number of UMIs not detected for a given gene. The full length fragmented reads without xxx’ enrichment or UMI were aligned using RSEM against a reference transcriptome  which is depleted of mtRNA  rRNA  and pseudogenes as described above. The command ‘rsem calculate expression’ was used with parameters  p 4   paired end   seed length 24 for performing the alignment. The resulting gene levels in FPKM was used to build the expression matrix. Genome build: GRCz10 dr81 Supplementary files format and content: Tab delimited text files include gene levels UMI counts or FKPM for each sample.,Embryonic cell,,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for 5’ enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,Fish of each genotype were incrossed and embryos were collected. Embryos were dechorionated and allowed to grow at 28°C in standard fish water containing methylene blue till 50% epiboly stage.,developmental stage:50% epiboly|library preparation batch:p2|genotype:Wild type TLAB,GSM2838423,GSM2838423: wt p2 S7; Danio rerio; RNA Seq,GSM2838423,,1,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for five prime enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina HiSeq 2500,,SRP123526,,,wt_p2_S7.R1.fastq.gz wt_p2_S7.R2.fastq.gz,fastq fastq,38703896.0,691141.0,GSM2838423 r1,0:24 1:32,A:10311865;C:8903162;G:8992515;T:10465321;N:31033,24,32,,,10311865,8903162,8992515,10465321,31033,SRX3357261,SRS2656342,SRA627905,GEO,"Schier Lab, Molecular and Cellular Biology, Harvard University",2,0.76715,0.8157,0.17322,0.18486,0.865,0.86665,0.6579,0.66292,24,32,B,B,biological fallback assumption,illumina,hiseq_era,full_length,poly_a,unknown,sc,single_cell_plate,smartseq,,United States,2017-11-02,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
44114,SRR6250336,SRX3357260,SRS2656341,SRP123526,PRJNA416939,Single cell RNAseq SMART seq2 of wild type TLAB and MZoep tz57 zebrafish embryos at 50% epiboly stage,GSE106466,Transcriptome Analysis,SMART seq2 was performed on single cells isolated from visually staged zebrafish embryos. Overall design: Samples were all sequenced in one batch. Some were generated with a five prime' UMI tagged method  and others are full length SMART seq2.,parent bioproject:PRJNA417291,pubmed:29700225,,wt p2 S6,GSM2838422,,tissue:Embryonic cell|developmental stage:50% epiboly|library preparation batch:p2|genotype:Wild type TLAB,wt p2 S6,Bases 1 8 the UMI and template switch were trimmed from the remaining reads  trimmed stretches of polyA or polyT longer than 6bp and removed any reads that were shorter than 15 bp post trimming. We identified and depleted mtRNA and rRNA reads from the remaining reads by using Bowtie2 to align all reads against a transcriptome of mtRNA and rRNA  using the parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed   no unal. Remaining reads were then aligned to a modified version of the Ensembl Zv10 release 81 zebrafish transcriptome using Bowtie2 with parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed. The transcriptome start site of each gene was padded by 100bp to increase alignment rates  since reads were generated from the 5’ end of the transcript  and not all transcripts had correct annotations for their 5’ end. Additionally  pseudogenes were removed from the transcriptome as they sometimes interfered with real genes  given our 5’ coordinate selection described below. We chose the best alignment per read as the read with the most 5’ alignment coordinate to a transcript among those that aligned to the correct strand and had an alignment score of no more than 4 below the best alignment score overall. We then iteratively collapsed pairs of UMIs  if the sum of the minimum Phred quality scores across non matching bases was at most 30 to account for mis matched bases that were likely due to sequencing errors. A collapsed pair was represented for future comparisons by the UMI with a larger prior number of aligned reads unless the other UMI had at least 75% as many reads aligned and had equal or better UMI quality scores for those alignments evaluated as the softmin over aligned reads of the softmax over quality scores for the UMI bases in an aligned read. post collapsing  the number of transcripts observed per gene was calculated. This value was then adjusted based on the probability of a collision the drawing at random of the same UMI multiple times  which was calculated from the number of UMIs not detected for a given gene. The full length fragmented reads without xxx’ enrichment or UMI were aligned using RSEM against a reference transcriptome  which is depleted of mtRNA  rRNA  and pseudogenes as described above. The command ‘rsem calculate expression’ was used with parameters  p 4   paired end   seed length 24 for performing the alignment. The resulting gene levels in FPKM was used to build the expression matrix. Genome build: GRCz10 dr81 Supplementary files format and content: Tab delimited text files include gene levels UMI counts or FKPM for each sample.,Embryonic cell,,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for 5’ enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,Fish of each genotype were incrossed and embryos were collected. Embryos were dechorionated and allowed to grow at 28°C in standard fish water containing methylene blue till 50% epiboly stage.,developmental stage:50% epiboly|library preparation batch:p2|genotype:Wild type TLAB,GSM2838422,GSM2838422: wt p2 S6; Danio rerio; RNA Seq,GSM2838422,,1,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for five prime enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina HiSeq 2500,,SRP123526,,,wt_p2_S6.R1.fastq.gz wt_p2_S6.R2.fastq.gz,fastq fastq,30542456.0,545401.0,GSM2838422 r1,0:24 1:32,A:8033321;C:7107172;G:7164454;T:8209836;N:27673,24,32,,,8033321,7107172,7164454,8209836,27673,SRX3357260,SRS2656341,SRA627905,GEO,"Schier Lab, Molecular and Cellular Biology, Harvard University",2,0.78373,0.83414,0.18981,0.2014,0.86235,0.86373,0.6968,0.7005,24,32,B,B,biological fallback assumption,illumina,hiseq_era,full_length,poly_a,unknown,sc,single_cell_plate,smartseq,,United States,2017-11-02,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
44115,SRR6250335,SRX3357259,SRS2656340,SRP123526,PRJNA416939,Single cell RNAseq SMART seq2 of wild type TLAB and MZoep tz57 zebrafish embryos at 50% epiboly stage,GSE106466,Transcriptome Analysis,SMART seq2 was performed on single cells isolated from visually staged zebrafish embryos. Overall design: Samples were all sequenced in one batch. Some were generated with a five prime' UMI tagged method  and others are full length SMART seq2.,parent bioproject:PRJNA417291,pubmed:29700225,,wt p2 S5,GSM2838421,,tissue:Embryonic cell|developmental stage:50% epiboly|library preparation batch:p2|genotype:Wild type TLAB,wt p2 S5,Bases 1 8 the UMI and template switch were trimmed from the remaining reads  trimmed stretches of polyA or polyT longer than 6bp and removed any reads that were shorter than 15 bp post trimming. We identified and depleted mtRNA and rRNA reads from the remaining reads by using Bowtie2 to align all reads against a transcriptome of mtRNA and rRNA  using the parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed   no unal. Remaining reads were then aligned to a modified version of the Ensembl Zv10 release 81 zebrafish transcriptome using Bowtie2 with parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed. The transcriptome start site of each gene was padded by 100bp to increase alignment rates  since reads were generated from the 5’ end of the transcript  and not all transcripts had correct annotations for their 5’ end. Additionally  pseudogenes were removed from the transcriptome as they sometimes interfered with real genes  given our 5’ coordinate selection described below. We chose the best alignment per read as the read with the most 5’ alignment coordinate to a transcript among those that aligned to the correct strand and had an alignment score of no more than 4 below the best alignment score overall. We then iteratively collapsed pairs of UMIs  if the sum of the minimum Phred quality scores across non matching bases was at most 30 to account for mis matched bases that were likely due to sequencing errors. A collapsed pair was represented for future comparisons by the UMI with a larger prior number of aligned reads unless the other UMI had at least 75% as many reads aligned and had equal or better UMI quality scores for those alignments evaluated as the softmin over aligned reads of the softmax over quality scores for the UMI bases in an aligned read. post collapsing  the number of transcripts observed per gene was calculated. This value was then adjusted based on the probability of a collision the drawing at random of the same UMI multiple times  which was calculated from the number of UMIs not detected for a given gene. The full length fragmented reads without xxx’ enrichment or UMI were aligned using RSEM against a reference transcriptome  which is depleted of mtRNA  rRNA  and pseudogenes as described above. The command ‘rsem calculate expression’ was used with parameters  p 4   paired end   seed length 24 for performing the alignment. The resulting gene levels in FPKM was used to build the expression matrix. Genome build: GRCz10 dr81 Supplementary files format and content: Tab delimited text files include gene levels UMI counts or FKPM for each sample.,Embryonic cell,,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for 5’ enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,Fish of each genotype were incrossed and embryos were collected. Embryos were dechorionated and allowed to grow at 28°C in standard fish water containing methylene blue till 50% epiboly stage.,developmental stage:50% epiboly|library preparation batch:p2|genotype:Wild type TLAB,GSM2838421,GSM2838421: wt p2 S5; Danio rerio; RNA Seq,GSM2838421,,1,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for five prime enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina HiSeq 2500,,SRP123526,,,wt_p2_S5.R1.fastq.gz wt_p2_S5.R2.fastq.gz,fastq fastq,34959568.0,624278.0,GSM2838421 r1,0:24 1:32,A:9180546;C:8129519;G:8220988;T:9399705;N:28810,24,32,,,9180546,8129519,8220988,9399705,28810,SRX3357259,SRS2656340,SRA627905,GEO,"Schier Lab, Molecular and Cellular Biology, Harvard University",2,0.78476,0.83524,0.21505,0.22989,0.86208,0.8636,0.67586,0.67373,24,32,B,B,biological fallback assumption,illumina,hiseq_era,full_length,poly_a,unknown,sc,single_cell_plate,smartseq,,United States,2017-11-02,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
44116,SRR6250334,SRX3357258,SRS2656339,SRP123526,PRJNA416939,Single cell RNAseq SMART seq2 of wild type TLAB and MZoep tz57 zebrafish embryos at 50% epiboly stage,GSE106466,Transcriptome Analysis,SMART seq2 was performed on single cells isolated from visually staged zebrafish embryos. Overall design: Samples were all sequenced in one batch. Some were generated with a five prime' UMI tagged method  and others are full length SMART seq2.,parent bioproject:PRJNA417291,pubmed:29700225,,wt p2 S4,GSM2838420,,tissue:Embryonic cell|developmental stage:50% epiboly|library preparation batch:p2|genotype:Wild type TLAB,wt p2 S4,Bases 1 8 the UMI and template switch were trimmed from the remaining reads  trimmed stretches of polyA or polyT longer than 6bp and removed any reads that were shorter than 15 bp post trimming. We identified and depleted mtRNA and rRNA reads from the remaining reads by using Bowtie2 to align all reads against a transcriptome of mtRNA and rRNA  using the parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed   no unal. Remaining reads were then aligned to a modified version of the Ensembl Zv10 release 81 zebrafish transcriptome using Bowtie2 with parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed. The transcriptome start site of each gene was padded by 100bp to increase alignment rates  since reads were generated from the 5’ end of the transcript  and not all transcripts had correct annotations for their 5’ end. Additionally  pseudogenes were removed from the transcriptome as they sometimes interfered with real genes  given our 5’ coordinate selection described below. We chose the best alignment per read as the read with the most 5’ alignment coordinate to a transcript among those that aligned to the correct strand and had an alignment score of no more than 4 below the best alignment score overall. We then iteratively collapsed pairs of UMIs  if the sum of the minimum Phred quality scores across non matching bases was at most 30 to account for mis matched bases that were likely due to sequencing errors. A collapsed pair was represented for future comparisons by the UMI with a larger prior number of aligned reads unless the other UMI had at least 75% as many reads aligned and had equal or better UMI quality scores for those alignments evaluated as the softmin over aligned reads of the softmax over quality scores for the UMI bases in an aligned read. post collapsing  the number of transcripts observed per gene was calculated. This value was then adjusted based on the probability of a collision the drawing at random of the same UMI multiple times  which was calculated from the number of UMIs not detected for a given gene. The full length fragmented reads without xxx’ enrichment or UMI were aligned using RSEM against a reference transcriptome  which is depleted of mtRNA  rRNA  and pseudogenes as described above. The command ‘rsem calculate expression’ was used with parameters  p 4   paired end   seed length 24 for performing the alignment. The resulting gene levels in FPKM was used to build the expression matrix. Genome build: GRCz10 dr81 Supplementary files format and content: Tab delimited text files include gene levels UMI counts or FKPM for each sample.,Embryonic cell,,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for 5’ enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,Fish of each genotype were incrossed and embryos were collected. Embryos were dechorionated and allowed to grow at 28°C in standard fish water containing methylene blue till 50% epiboly stage.,developmental stage:50% epiboly|library preparation batch:p2|genotype:Wild type TLAB,GSM2838420,GSM2838420: wt p2 S4; Danio rerio; RNA Seq,GSM2838420,,1,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for five prime enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina HiSeq 2500,,SRP123526,,,wt_p2_S4.R1.fastq.gz wt_p2_S4.R2.fastq.gz,fastq fastq,11573184.0,206664.0,GSM2838420 r1,0:24 1:32,A:3099219;C:2620762;G:2639101;T:3203925;N:10177,24,32,,,3099219,2620762,2639101,3203925,10177,SRX3357258,SRS2656339,SRA627905,GEO,"Schier Lab, Molecular and Cellular Biology, Harvard University",2,0.70775,0.76438,0.31372,0.33513,0.91831,0.92048,0.71842,0.71709,24,32,B,B,biological fallback assumption,illumina,hiseq_era,full_length,poly_a,unknown,sc,single_cell_plate,smartseq,,United States,2017-11-02,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
44117,SRR6250333,SRX3357256,SRS2656338,SRP123526,PRJNA416939,Single cell RNAseq SMART seq2 of wild type TLAB and MZoep tz57 zebrafish embryos at 50% epiboly stage,GSE106466,Transcriptome Analysis,SMART seq2 was performed on single cells isolated from visually staged zebrafish embryos. Overall design: Samples were all sequenced in one batch. Some were generated with a five prime' UMI tagged method  and others are full length SMART seq2.,parent bioproject:PRJNA417291,pubmed:29700225,,wt p2 S3,GSM2838419,,tissue:Embryonic cell|developmental stage:50% epiboly|library preparation batch:p2|genotype:Wild type TLAB,wt p2 S3,Bases 1 8 the UMI and template switch were trimmed from the remaining reads  trimmed stretches of polyA or polyT longer than 6bp and removed any reads that were shorter than 15 bp post trimming. We identified and depleted mtRNA and rRNA reads from the remaining reads by using Bowtie2 to align all reads against a transcriptome of mtRNA and rRNA  using the parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed   no unal. Remaining reads were then aligned to a modified version of the Ensembl Zv10 release 81 zebrafish transcriptome using Bowtie2 with parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed. The transcriptome start site of each gene was padded by 100bp to increase alignment rates  since reads were generated from the 5’ end of the transcript  and not all transcripts had correct annotations for their 5’ end. Additionally  pseudogenes were removed from the transcriptome as they sometimes interfered with real genes  given our 5’ coordinate selection described below. We chose the best alignment per read as the read with the most 5’ alignment coordinate to a transcript among those that aligned to the correct strand and had an alignment score of no more than 4 below the best alignment score overall. We then iteratively collapsed pairs of UMIs  if the sum of the minimum Phred quality scores across non matching bases was at most 30 to account for mis matched bases that were likely due to sequencing errors. A collapsed pair was represented for future comparisons by the UMI with a larger prior number of aligned reads unless the other UMI had at least 75% as many reads aligned and had equal or better UMI quality scores for those alignments evaluated as the softmin over aligned reads of the softmax over quality scores for the UMI bases in an aligned read. post collapsing  the number of transcripts observed per gene was calculated. This value was then adjusted based on the probability of a collision the drawing at random of the same UMI multiple times  which was calculated from the number of UMIs not detected for a given gene. The full length fragmented reads without xxx’ enrichment or UMI were aligned using RSEM against a reference transcriptome  which is depleted of mtRNA  rRNA  and pseudogenes as described above. The command ‘rsem calculate expression’ was used with parameters  p 4   paired end   seed length 24 for performing the alignment. The resulting gene levels in FPKM was used to build the expression matrix. Genome build: GRCz10 dr81 Supplementary files format and content: Tab delimited text files include gene levels UMI counts or FKPM for each sample.,Embryonic cell,,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for 5’ enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,Fish of each genotype were incrossed and embryos were collected. Embryos were dechorionated and allowed to grow at 28°C in standard fish water containing methylene blue till 50% epiboly stage.,developmental stage:50% epiboly|library preparation batch:p2|genotype:Wild type TLAB,GSM2838419,GSM2838419: wt p2 S3; Danio rerio; RNA Seq,GSM2838419,,1,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for five prime enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina HiSeq 2500,,SRP123526,,,wt_p2_S3.R1.fastq.gz wt_p2_S3.R2.fastq.gz,fastq fastq,37763040.0,674340.0,GSM2838419 r1,0:24 1:32,A:9993306;C:8699070;G:8817530;T:10222682;N:30452,24,32,,,9993306,8699070,8817530,10222682,30452,SRX3357256,SRS2656338,SRA627905,GEO,"Schier Lab, Molecular and Cellular Biology, Harvard University",2,0.7863,0.83568,0.22435,0.23906,0.84331,0.84228,0.35971,0.66377,24,32,B,B,biological fallback assumption,illumina,hiseq_era,full_length,poly_a,unknown,sc,single_cell_plate,smartseq,,United States,2017-11-02,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
44118,SRR6250332,SRX3357255,SRS2656336,SRP123526,PRJNA416939,Single cell RNAseq SMART seq2 of wild type TLAB and MZoep tz57 zebrafish embryos at 50% epiboly stage,GSE106466,Transcriptome Analysis,SMART seq2 was performed on single cells isolated from visually staged zebrafish embryos. Overall design: Samples were all sequenced in one batch. Some were generated with a five prime' UMI tagged method  and others are full length SMART seq2.,parent bioproject:PRJNA417291,pubmed:29700225,,wt p2 S2,GSM2838418,,tissue:Embryonic cell|developmental stage:50% epiboly|library preparation batch:p2|genotype:Wild type TLAB,wt p2 S2,Bases 1 8 the UMI and template switch were trimmed from the remaining reads  trimmed stretches of polyA or polyT longer than 6bp and removed any reads that were shorter than 15 bp post trimming. We identified and depleted mtRNA and rRNA reads from the remaining reads by using Bowtie2 to align all reads against a transcriptome of mtRNA and rRNA  using the parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed   no unal. Remaining reads were then aligned to a modified version of the Ensembl Zv10 release 81 zebrafish transcriptome using Bowtie2 with parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed. The transcriptome start site of each gene was padded by 100bp to increase alignment rates  since reads were generated from the 5’ end of the transcript  and not all transcripts had correct annotations for their 5’ end. Additionally  pseudogenes were removed from the transcriptome as they sometimes interfered with real genes  given our 5’ coordinate selection described below. We chose the best alignment per read as the read with the most 5’ alignment coordinate to a transcript among those that aligned to the correct strand and had an alignment score of no more than 4 below the best alignment score overall. We then iteratively collapsed pairs of UMIs  if the sum of the minimum Phred quality scores across non matching bases was at most 30 to account for mis matched bases that were likely due to sequencing errors. A collapsed pair was represented for future comparisons by the UMI with a larger prior number of aligned reads unless the other UMI had at least 75% as many reads aligned and had equal or better UMI quality scores for those alignments evaluated as the softmin over aligned reads of the softmax over quality scores for the UMI bases in an aligned read. post collapsing  the number of transcripts observed per gene was calculated. This value was then adjusted based on the probability of a collision the drawing at random of the same UMI multiple times  which was calculated from the number of UMIs not detected for a given gene. The full length fragmented reads without xxx’ enrichment or UMI were aligned using RSEM against a reference transcriptome  which is depleted of mtRNA  rRNA  and pseudogenes as described above. The command ‘rsem calculate expression’ was used with parameters  p 4   paired end   seed length 24 for performing the alignment. The resulting gene levels in FPKM was used to build the expression matrix. Genome build: GRCz10 dr81 Supplementary files format and content: Tab delimited text files include gene levels UMI counts or FKPM for each sample.,Embryonic cell,,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for 5’ enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,Fish of each genotype were incrossed and embryos were collected. Embryos were dechorionated and allowed to grow at 28°C in standard fish water containing methylene blue till 50% epiboly stage.,developmental stage:50% epiboly|library preparation batch:p2|genotype:Wild type TLAB,GSM2838418,GSM2838418: wt p2 S2; Danio rerio; RNA Seq,GSM2838418,,1,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for five prime enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina HiSeq 2500,,SRP123526,,,wt_p2_S2.R1.fastq.gz wt_p2_S2.R2.fastq.gz,fastq fastq,32485040.0,580090.0,GSM2838418 r1,0:24 1:32,A:8505580;C:7579715;G:7669107;T:8703486;N:27152,24,32,,,8505580,7579715,7669107,8703486,27152,SRX3357255,SRS2656336,SRA627905,GEO,"Schier Lab, Molecular and Cellular Biology, Harvard University",2,0.7719,0.82478,0.20255,0.21629,0.85234,0.85313,0.65578,0.64124,24,32,B,B,biological fallback assumption,illumina,hiseq_era,full_length,poly_a,unknown,sc,single_cell_plate,smartseq,,United States,2017-11-02,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
44119,SRR6250331,SRX3357254,SRS2656334,SRP123526,PRJNA416939,Single cell RNAseq SMART seq2 of wild type TLAB and MZoep tz57 zebrafish embryos at 50% epiboly stage,GSE106466,Transcriptome Analysis,SMART seq2 was performed on single cells isolated from visually staged zebrafish embryos. Overall design: Samples were all sequenced in one batch. Some were generated with a five prime' UMI tagged method  and others are full length SMART seq2.,parent bioproject:PRJNA417291,pubmed:29700225,,wt p2 S1,GSM2838417,,tissue:Embryonic cell|developmental stage:50% epiboly|library preparation batch:p2|genotype:Wild type TLAB,wt p2 S1,Bases 1 8 the UMI and template switch were trimmed from the remaining reads  trimmed stretches of polyA or polyT longer than 6bp and removed any reads that were shorter than 15 bp post trimming. We identified and depleted mtRNA and rRNA reads from the remaining reads by using Bowtie2 to align all reads against a transcriptome of mtRNA and rRNA  using the parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed   no unal. Remaining reads were then aligned to a modified version of the Ensembl Zv10 release 81 zebrafish transcriptome using Bowtie2 with parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed. The transcriptome start site of each gene was padded by 100bp to increase alignment rates  since reads were generated from the 5’ end of the transcript  and not all transcripts had correct annotations for their 5’ end. Additionally  pseudogenes were removed from the transcriptome as they sometimes interfered with real genes  given our 5’ coordinate selection described below. We chose the best alignment per read as the read with the most 5’ alignment coordinate to a transcript among those that aligned to the correct strand and had an alignment score of no more than 4 below the best alignment score overall. We then iteratively collapsed pairs of UMIs  if the sum of the minimum Phred quality scores across non matching bases was at most 30 to account for mis matched bases that were likely due to sequencing errors. A collapsed pair was represented for future comparisons by the UMI with a larger prior number of aligned reads unless the other UMI had at least 75% as many reads aligned and had equal or better UMI quality scores for those alignments evaluated as the softmin over aligned reads of the softmax over quality scores for the UMI bases in an aligned read. post collapsing  the number of transcripts observed per gene was calculated. This value was then adjusted based on the probability of a collision the drawing at random of the same UMI multiple times  which was calculated from the number of UMIs not detected for a given gene. The full length fragmented reads without xxx’ enrichment or UMI were aligned using RSEM against a reference transcriptome  which is depleted of mtRNA  rRNA  and pseudogenes as described above. The command ‘rsem calculate expression’ was used with parameters  p 4   paired end   seed length 24 for performing the alignment. The resulting gene levels in FPKM was used to build the expression matrix. Genome build: GRCz10 dr81 Supplementary files format and content: Tab delimited text files include gene levels UMI counts or FKPM for each sample.,Embryonic cell,,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for 5’ enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,Fish of each genotype were incrossed and embryos were collected. Embryos were dechorionated and allowed to grow at 28°C in standard fish water containing methylene blue till 50% epiboly stage.,developmental stage:50% epiboly|library preparation batch:p2|genotype:Wild type TLAB,GSM2838417,GSM2838417: wt p2 S1; Danio rerio; RNA Seq,GSM2838417,,1,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for five prime enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina HiSeq 2500,,SRP123526,,,wt_p2_S1.R1.fastq.gz wt_p2_S1.R2.fastq.gz,fastq fastq,36153488.0,645598.0,GSM2838417 r1,0:24 1:32,A:9482693;C:8433376;G:8529260;T:9678247;N:29912,24,32,,,9482693,8433376,8529260,9678247,29912,SRX3357254,SRS2656334,SRA627905,GEO,"Schier Lab, Molecular and Cellular Biology, Harvard University",2,0.7857,0.83959,0.20506,0.22028,0.85102,0.85121,0.68415,0.69308,24,32,B,B,biological fallback assumption,illumina,hiseq_era,full_length,poly_a,unknown,sc,single_cell_plate,smartseq,,United States,2017-11-02,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
44120,SRR6250330,SRX3357253,SRS2656335,SRP123526,PRJNA416939,Single cell RNAseq SMART seq2 of wild type TLAB and MZoep tz57 zebrafish embryos at 50% epiboly stage,GSE106466,Transcriptome Analysis,SMART seq2 was performed on single cells isolated from visually staged zebrafish embryos. Overall design: Samples were all sequenced in one batch. Some were generated with a five prime' UMI tagged method  and others are full length SMART seq2.,parent bioproject:PRJNA417291,pubmed:29700225,,wt p2 S14,GSM2838416,,tissue:Embryonic cell|developmental stage:50% epiboly|library preparation batch:p2|genotype:Wild type TLAB,wt p2 S14,Bases 1 8 the UMI and template switch were trimmed from the remaining reads  trimmed stretches of polyA or polyT longer than 6bp and removed any reads that were shorter than 15 bp post trimming. We identified and depleted mtRNA and rRNA reads from the remaining reads by using Bowtie2 to align all reads against a transcriptome of mtRNA and rRNA  using the parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed   no unal. Remaining reads were then aligned to a modified version of the Ensembl Zv10 release 81 zebrafish transcriptome using Bowtie2 with parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed. The transcriptome start site of each gene was padded by 100bp to increase alignment rates  since reads were generated from the 5’ end of the transcript  and not all transcripts had correct annotations for their 5’ end. Additionally  pseudogenes were removed from the transcriptome as they sometimes interfered with real genes  given our 5’ coordinate selection described below. We chose the best alignment per read as the read with the most 5’ alignment coordinate to a transcript among those that aligned to the correct strand and had an alignment score of no more than 4 below the best alignment score overall. We then iteratively collapsed pairs of UMIs  if the sum of the minimum Phred quality scores across non matching bases was at most 30 to account for mis matched bases that were likely due to sequencing errors. A collapsed pair was represented for future comparisons by the UMI with a larger prior number of aligned reads unless the other UMI had at least 75% as many reads aligned and had equal or better UMI quality scores for those alignments evaluated as the softmin over aligned reads of the softmax over quality scores for the UMI bases in an aligned read. post collapsing  the number of transcripts observed per gene was calculated. This value was then adjusted based on the probability of a collision the drawing at random of the same UMI multiple times  which was calculated from the number of UMIs not detected for a given gene. The full length fragmented reads without xxx’ enrichment or UMI were aligned using RSEM against a reference transcriptome  which is depleted of mtRNA  rRNA  and pseudogenes as described above. The command ‘rsem calculate expression’ was used with parameters  p 4   paired end   seed length 24 for performing the alignment. The resulting gene levels in FPKM was used to build the expression matrix. Genome build: GRCz10 dr81 Supplementary files format and content: Tab delimited text files include gene levels UMI counts or FKPM for each sample.,Embryonic cell,,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for 5’ enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,Fish of each genotype were incrossed and embryos were collected. Embryos were dechorionated and allowed to grow at 28°C in standard fish water containing methylene blue till 50% epiboly stage.,developmental stage:50% epiboly|library preparation batch:p2|genotype:Wild type TLAB,GSM2838416,GSM2838416: wt p2 S14; Danio rerio; RNA Seq,GSM2838416,,1,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for five prime enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina HiSeq 2500,,SRP123526,,,wt_p2_S14.R1.fastq.gz wt_p2_S14.R2.fastq.gz,fastq fastq,35662032.0,636822.0,GSM2838416 r1,0:24 1:32,A:9220957;C:8441374;G:8533782;T:9437149;N:28770,24,32,,,9220957,8441374,8533782,9437149,28770,SRX3357253,SRS2656335,SRA627905,GEO,"Schier Lab, Molecular and Cellular Biology, Harvard University",2,0.78999,0.8417,0.21214,0.22792,0.85397,0.85421,0.68135,0.68129,24,32,B,B,biological fallback assumption,illumina,hiseq_era,full_length,poly_a,unknown,sc,single_cell_plate,smartseq,,United States,2017-11-02,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
44121,SRR6250329,SRX3357252,SRS2656333,SRP123526,PRJNA416939,Single cell RNAseq SMART seq2 of wild type TLAB and MZoep tz57 zebrafish embryos at 50% epiboly stage,GSE106466,Transcriptome Analysis,SMART seq2 was performed on single cells isolated from visually staged zebrafish embryos. Overall design: Samples were all sequenced in one batch. Some were generated with a five prime' UMI tagged method  and others are full length SMART seq2.,parent bioproject:PRJNA417291,pubmed:29700225,,wt p2 S13,GSM2838415,,tissue:Embryonic cell|developmental stage:50% epiboly|library preparation batch:p2|genotype:Wild type TLAB,wt p2 S13,Bases 1 8 the UMI and template switch were trimmed from the remaining reads  trimmed stretches of polyA or polyT longer than 6bp and removed any reads that were shorter than 15 bp post trimming. We identified and depleted mtRNA and rRNA reads from the remaining reads by using Bowtie2 to align all reads against a transcriptome of mtRNA and rRNA  using the parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed   no unal. Remaining reads were then aligned to a modified version of the Ensembl Zv10 release 81 zebrafish transcriptome using Bowtie2 with parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed. The transcriptome start site of each gene was padded by 100bp to increase alignment rates  since reads were generated from the 5’ end of the transcript  and not all transcripts had correct annotations for their 5’ end. Additionally  pseudogenes were removed from the transcriptome as they sometimes interfered with real genes  given our 5’ coordinate selection described below. We chose the best alignment per read as the read with the most 5’ alignment coordinate to a transcript among those that aligned to the correct strand and had an alignment score of no more than 4 below the best alignment score overall. We then iteratively collapsed pairs of UMIs  if the sum of the minimum Phred quality scores across non matching bases was at most 30 to account for mis matched bases that were likely due to sequencing errors. A collapsed pair was represented for future comparisons by the UMI with a larger prior number of aligned reads unless the other UMI had at least 75% as many reads aligned and had equal or better UMI quality scores for those alignments evaluated as the softmin over aligned reads of the softmax over quality scores for the UMI bases in an aligned read. post collapsing  the number of transcripts observed per gene was calculated. This value was then adjusted based on the probability of a collision the drawing at random of the same UMI multiple times  which was calculated from the number of UMIs not detected for a given gene. The full length fragmented reads without xxx’ enrichment or UMI were aligned using RSEM against a reference transcriptome  which is depleted of mtRNA  rRNA  and pseudogenes as described above. The command ‘rsem calculate expression’ was used with parameters  p 4   paired end   seed length 24 for performing the alignment. The resulting gene levels in FPKM was used to build the expression matrix. Genome build: GRCz10 dr81 Supplementary files format and content: Tab delimited text files include gene levels UMI counts or FKPM for each sample.,Embryonic cell,,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for 5’ enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,Fish of each genotype were incrossed and embryos were collected. Embryos were dechorionated and allowed to grow at 28°C in standard fish water containing methylene blue till 50% epiboly stage.,developmental stage:50% epiboly|library preparation batch:p2|genotype:Wild type TLAB,GSM2838415,GSM2838415: wt p2 S13; Danio rerio; RNA Seq,GSM2838415,,1,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for five prime enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina HiSeq 2500,,SRP123526,,,wt_p2_S13.R1.fastq.gz wt_p2_S13.R2.fastq.gz,fastq fastq,35179312.0,628202.0,GSM2838415 r1,0:24 1:32,A:9165054;C:8266293;G:8343113;T:9376252;N:28600,24,32,,,9165054,8266293,8343113,9376252,28600,SRX3357252,SRS2656333,SRA627905,GEO,"Schier Lab, Molecular and Cellular Biology, Harvard University",2,0.78329,0.83757,0.22073,0.23652,0.87048,0.87054,0.68987,0.69553,24,32,B,B,biological fallback assumption,illumina,hiseq_era,full_length,poly_a,unknown,sc,single_cell_plate,smartseq,,United States,2017-11-02,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
44122,SRR6250328,SRX3357251,SRS2656331,SRP123526,PRJNA416939,Single cell RNAseq SMART seq2 of wild type TLAB and MZoep tz57 zebrafish embryos at 50% epiboly stage,GSE106466,Transcriptome Analysis,SMART seq2 was performed on single cells isolated from visually staged zebrafish embryos. Overall design: Samples were all sequenced in one batch. Some were generated with a five prime' UMI tagged method  and others are full length SMART seq2.,parent bioproject:PRJNA417291,pubmed:29700225,,wt p2 S12,GSM2838414,,tissue:Embryonic cell|developmental stage:50% epiboly|library preparation batch:p2|genotype:Wild type TLAB,wt p2 S12,Bases 1 8 the UMI and template switch were trimmed from the remaining reads  trimmed stretches of polyA or polyT longer than 6bp and removed any reads that were shorter than 15 bp post trimming. We identified and depleted mtRNA and rRNA reads from the remaining reads by using Bowtie2 to align all reads against a transcriptome of mtRNA and rRNA  using the parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed   no unal. Remaining reads were then aligned to a modified version of the Ensembl Zv10 release 81 zebrafish transcriptome using Bowtie2 with parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed. The transcriptome start site of each gene was padded by 100bp to increase alignment rates  since reads were generated from the 5’ end of the transcript  and not all transcripts had correct annotations for their 5’ end. Additionally  pseudogenes were removed from the transcriptome as they sometimes interfered with real genes  given our 5’ coordinate selection described below. We chose the best alignment per read as the read with the most 5’ alignment coordinate to a transcript among those that aligned to the correct strand and had an alignment score of no more than 4 below the best alignment score overall. We then iteratively collapsed pairs of UMIs  if the sum of the minimum Phred quality scores across non matching bases was at most 30 to account for mis matched bases that were likely due to sequencing errors. A collapsed pair was represented for future comparisons by the UMI with a larger prior number of aligned reads unless the other UMI had at least 75% as many reads aligned and had equal or better UMI quality scores for those alignments evaluated as the softmin over aligned reads of the softmax over quality scores for the UMI bases in an aligned read. post collapsing  the number of transcripts observed per gene was calculated. This value was then adjusted based on the probability of a collision the drawing at random of the same UMI multiple times  which was calculated from the number of UMIs not detected for a given gene. The full length fragmented reads without xxx’ enrichment or UMI were aligned using RSEM against a reference transcriptome  which is depleted of mtRNA  rRNA  and pseudogenes as described above. The command ‘rsem calculate expression’ was used with parameters  p 4   paired end   seed length 24 for performing the alignment. The resulting gene levels in FPKM was used to build the expression matrix. Genome build: GRCz10 dr81 Supplementary files format and content: Tab delimited text files include gene levels UMI counts or FKPM for each sample.,Embryonic cell,,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for 5’ enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,Fish of each genotype were incrossed and embryos were collected. Embryos were dechorionated and allowed to grow at 28°C in standard fish water containing methylene blue till 50% epiboly stage.,developmental stage:50% epiboly|library preparation batch:p2|genotype:Wild type TLAB,GSM2838414,GSM2838414: wt p2 S12; Danio rerio; RNA Seq,GSM2838414,,1,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for five prime enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina HiSeq 2500,,SRP123526,,,wt_p2_S12.R1.fastq.gz wt_p2_S12.R2.fastq.gz,fastq fastq,39207392.0,700132.0,GSM2838414 r1,0:24 1:32,A:10192020;C:9230236;G:9318388;T:10435252;N:31496,24,32,,,10192020,9230236,9318388,10435252,31496,SRX3357251,SRS2656331,SRA627905,GEO,"Schier Lab, Molecular and Cellular Biology, Harvard University",2,0.78483,0.84016,0.22133,0.23855,0.8743,0.8757,0.70284,0.70386,24,32,B,B,biological fallback assumption,illumina,hiseq_era,full_length,poly_a,unknown,sc,single_cell_plate,smartseq,,United States,2017-11-02,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
44123,SRR6250327,SRX3357250,SRS2656332,SRP123526,PRJNA416939,Single cell RNAseq SMART seq2 of wild type TLAB and MZoep tz57 zebrafish embryos at 50% epiboly stage,GSE106466,Transcriptome Analysis,SMART seq2 was performed on single cells isolated from visually staged zebrafish embryos. Overall design: Samples were all sequenced in one batch. Some were generated with a five prime' UMI tagged method  and others are full length SMART seq2.,parent bioproject:PRJNA417291,pubmed:29700225,,wt p2 S11,GSM2838413,,tissue:Embryonic cell|developmental stage:50% epiboly|library preparation batch:p2|genotype:Wild type TLAB,wt p2 S11,Bases 1 8 the UMI and template switch were trimmed from the remaining reads  trimmed stretches of polyA or polyT longer than 6bp and removed any reads that were shorter than 15 bp post trimming. We identified and depleted mtRNA and rRNA reads from the remaining reads by using Bowtie2 to align all reads against a transcriptome of mtRNA and rRNA  using the parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed   no unal. Remaining reads were then aligned to a modified version of the Ensembl Zv10 release 81 zebrafish transcriptome using Bowtie2 with parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed. The transcriptome start site of each gene was padded by 100bp to increase alignment rates  since reads were generated from the 5’ end of the transcript  and not all transcripts had correct annotations for their 5’ end. Additionally  pseudogenes were removed from the transcriptome as they sometimes interfered with real genes  given our 5’ coordinate selection described below. We chose the best alignment per read as the read with the most 5’ alignment coordinate to a transcript among those that aligned to the correct strand and had an alignment score of no more than 4 below the best alignment score overall. We then iteratively collapsed pairs of UMIs  if the sum of the minimum Phred quality scores across non matching bases was at most 30 to account for mis matched bases that were likely due to sequencing errors. A collapsed pair was represented for future comparisons by the UMI with a larger prior number of aligned reads unless the other UMI had at least 75% as many reads aligned and had equal or better UMI quality scores for those alignments evaluated as the softmin over aligned reads of the softmax over quality scores for the UMI bases in an aligned read. post collapsing  the number of transcripts observed per gene was calculated. This value was then adjusted based on the probability of a collision the drawing at random of the same UMI multiple times  which was calculated from the number of UMIs not detected for a given gene. The full length fragmented reads without xxx’ enrichment or UMI were aligned using RSEM against a reference transcriptome  which is depleted of mtRNA  rRNA  and pseudogenes as described above. The command ‘rsem calculate expression’ was used with parameters  p 4   paired end   seed length 24 for performing the alignment. The resulting gene levels in FPKM was used to build the expression matrix. Genome build: GRCz10 dr81 Supplementary files format and content: Tab delimited text files include gene levels UMI counts or FKPM for each sample.,Embryonic cell,,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for 5’ enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,Fish of each genotype were incrossed and embryos were collected. Embryos were dechorionated and allowed to grow at 28°C in standard fish water containing methylene blue till 50% epiboly stage.,developmental stage:50% epiboly|library preparation batch:p2|genotype:Wild type TLAB,GSM2838413,GSM2838413: wt p2 S11; Danio rerio; RNA Seq,GSM2838413,,1,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for five prime enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina HiSeq 2500,,SRP123526,,,wt_p2_S11.R1.fastq.gz wt_p2_S11.R2.fastq.gz,fastq fastq,40102944.0,716124.0,GSM2838413 r1,0:24 1:32,A:10511641;C:9360932;G:9463806;T:10731533;N:35032,24,32,,,10511641,9360932,9463806,10731533,35032,SRX3357250,SRS2656332,SRA627905,GEO,"Schier Lab, Molecular and Cellular Biology, Harvard University",2,0.78698,0.83686,0.21254,0.22573,0.85985,0.86011,0.67858,0.67878,24,32,B,B,biological fallback assumption,illumina,hiseq_era,full_length,poly_a,unknown,sc,single_cell_plate,smartseq,,United States,2017-11-02,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
44124,SRR6250326,SRX3357249,SRS2656330,SRP123526,PRJNA416939,Single cell RNAseq SMART seq2 of wild type TLAB and MZoep tz57 zebrafish embryos at 50% epiboly stage,GSE106466,Transcriptome Analysis,SMART seq2 was performed on single cells isolated from visually staged zebrafish embryos. Overall design: Samples were all sequenced in one batch. Some were generated with a five prime' UMI tagged method  and others are full length SMART seq2.,parent bioproject:PRJNA417291,pubmed:29700225,,wt p2 S10,GSM2838412,,tissue:Embryonic cell|developmental stage:50% epiboly|library preparation batch:p2|genotype:Wild type TLAB,wt p2 S10,Bases 1 8 the UMI and template switch were trimmed from the remaining reads  trimmed stretches of polyA or polyT longer than 6bp and removed any reads that were shorter than 15 bp post trimming. We identified and depleted mtRNA and rRNA reads from the remaining reads by using Bowtie2 to align all reads against a transcriptome of mtRNA and rRNA  using the parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed   no unal. Remaining reads were then aligned to a modified version of the Ensembl Zv10 release 81 zebrafish transcriptome using Bowtie2 with parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed. The transcriptome start site of each gene was padded by 100bp to increase alignment rates  since reads were generated from the 5’ end of the transcript  and not all transcripts had correct annotations for their 5’ end. Additionally  pseudogenes were removed from the transcriptome as they sometimes interfered with real genes  given our 5’ coordinate selection described below. We chose the best alignment per read as the read with the most 5’ alignment coordinate to a transcript among those that aligned to the correct strand and had an alignment score of no more than 4 below the best alignment score overall. We then iteratively collapsed pairs of UMIs  if the sum of the minimum Phred quality scores across non matching bases was at most 30 to account for mis matched bases that were likely due to sequencing errors. A collapsed pair was represented for future comparisons by the UMI with a larger prior number of aligned reads unless the other UMI had at least 75% as many reads aligned and had equal or better UMI quality scores for those alignments evaluated as the softmin over aligned reads of the softmax over quality scores for the UMI bases in an aligned read. post collapsing  the number of transcripts observed per gene was calculated. This value was then adjusted based on the probability of a collision the drawing at random of the same UMI multiple times  which was calculated from the number of UMIs not detected for a given gene. The full length fragmented reads without xxx’ enrichment or UMI were aligned using RSEM against a reference transcriptome  which is depleted of mtRNA  rRNA  and pseudogenes as described above. The command ‘rsem calculate expression’ was used with parameters  p 4   paired end   seed length 24 for performing the alignment. The resulting gene levels in FPKM was used to build the expression matrix. Genome build: GRCz10 dr81 Supplementary files format and content: Tab delimited text files include gene levels UMI counts or FKPM for each sample.,Embryonic cell,,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for 5’ enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,Fish of each genotype were incrossed and embryos were collected. Embryos were dechorionated and allowed to grow at 28°C in standard fish water containing methylene blue till 50% epiboly stage.,developmental stage:50% epiboly|library preparation batch:p2|genotype:Wild type TLAB,GSM2838412,GSM2838412: wt p2 S10; Danio rerio; RNA Seq,GSM2838412,,1,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for five prime enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina HiSeq 2500,,SRP123526,,,wt_p2_S10.R1.fastq.gz wt_p2_S10.R2.fastq.gz,fastq fastq,44972536.0,803081.0,GSM2838412 r1,0:24 1:32,A:11615173;C:10671699;G:10744448;T:11903998;N:37218,24,32,,,11615173,10671699,10744448,11903998,37218,SRX3357249,SRS2656330,SRA627905,GEO,"Schier Lab, Molecular and Cellular Biology, Harvard University",2,0.78871,0.84338,0.21928,0.23653,0.8574,0.85772,0.71222,0.69441,24,32,B,B,biological fallback assumption,illumina,hiseq_era,full_length,poly_a,unknown,sc,single_cell_plate,smartseq,,United States,2017-11-02,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
44125,SRR6250325,SRX3357248,SRS2656329,SRP123526,PRJNA416939,Single cell RNAseq SMART seq2 of wild type TLAB and MZoep tz57 zebrafish embryos at 50% epiboly stage,GSE106466,Transcriptome Analysis,SMART seq2 was performed on single cells isolated from visually staged zebrafish embryos. Overall design: Samples were all sequenced in one batch. Some were generated with a five prime' UMI tagged method  and others are full length SMART seq2.,parent bioproject:PRJNA417291,pubmed:29700225,,wt p1 S3,GSM2838411,,tissue:Embryonic cell|developmental stage:50% epiboly|library preparation batch:p1|genotype:Wild type TLAB,wt p1 S3,Bases 1 8 the UMI and template switch were trimmed from the remaining reads  trimmed stretches of polyA or polyT longer than 6bp and removed any reads that were shorter than 15 bp post trimming. We identified and depleted mtRNA and rRNA reads from the remaining reads by using Bowtie2 to align all reads against a transcriptome of mtRNA and rRNA  using the parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed   no unal. Remaining reads were then aligned to a modified version of the Ensembl Zv10 release 81 zebrafish transcriptome using Bowtie2 with parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed. The transcriptome start site of each gene was padded by 100bp to increase alignment rates  since reads were generated from the 5’ end of the transcript  and not all transcripts had correct annotations for their 5’ end. Additionally  pseudogenes were removed from the transcriptome as they sometimes interfered with real genes  given our 5’ coordinate selection described below. We chose the best alignment per read as the read with the most 5’ alignment coordinate to a transcript among those that aligned to the correct strand and had an alignment score of no more than 4 below the best alignment score overall. We then iteratively collapsed pairs of UMIs  if the sum of the minimum Phred quality scores across non matching bases was at most 30 to account for mis matched bases that were likely due to sequencing errors. A collapsed pair was represented for future comparisons by the UMI with a larger prior number of aligned reads unless the other UMI had at least 75% as many reads aligned and had equal or better UMI quality scores for those alignments evaluated as the softmin over aligned reads of the softmax over quality scores for the UMI bases in an aligned read. post collapsing  the number of transcripts observed per gene was calculated. This value was then adjusted based on the probability of a collision the drawing at random of the same UMI multiple times  which was calculated from the number of UMIs not detected for a given gene. The full length fragmented reads without xxx’ enrichment or UMI were aligned using RSEM against a reference transcriptome  which is depleted of mtRNA  rRNA  and pseudogenes as described above. The command ‘rsem calculate expression’ was used with parameters  p 4   paired end   seed length 24 for performing the alignment. The resulting gene levels in FPKM was used to build the expression matrix. Genome build: GRCz10 dr81 Supplementary files format and content: Tab delimited text files include gene levels UMI counts or FKPM for each sample.,Embryonic cell,,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for 5’ enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,Fish of each genotype were incrossed and embryos were collected. Embryos were dechorionated and allowed to grow at 28°C in standard fish water containing methylene blue till 50% epiboly stage.,developmental stage:50% epiboly|library preparation batch:p1|genotype:Wild type TLAB,GSM2838411,GSM2838411: wt p1 S3; Danio rerio; RNA Seq,GSM2838411,,1,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for five prime enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina HiSeq 2500,,SRP123526,,,wt_p1_S3.R1.fastq.gz wt_p1_S3.R2.fastq.gz,fastq fastq,95933432.0,1713097.0,GSM2838411 r1,0:24 1:32,A:26332108;C:21430855;G:21729899;T:26360908;N:79662,24,32,,,26332108,21430855,21729899,26360908,79662,SRX3357248,SRS2656329,SRA627905,GEO,"Schier Lab, Molecular and Cellular Biology, Harvard University",2,0.78047,0.82824,0.10747,0.11108,0.86304,0.86352,0.73589,0.73908,24,32,B,B,biological fallback assumption,illumina,hiseq_era,full_length,poly_a,unknown,sc,single_cell_plate,smartseq,,United States,2017-11-02,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
44126,SRR6250324,SRX3357247,SRS2656328,SRP123526,PRJNA416939,Single cell RNAseq SMART seq2 of wild type TLAB and MZoep tz57 zebrafish embryos at 50% epiboly stage,GSE106466,Transcriptome Analysis,SMART seq2 was performed on single cells isolated from visually staged zebrafish embryos. Overall design: Samples were all sequenced in one batch. Some were generated with a five prime' UMI tagged method  and others are full length SMART seq2.,parent bioproject:PRJNA417291,pubmed:29700225,,wt p1 S2,GSM2838410,,tissue:Embryonic cell|developmental stage:50% epiboly|library preparation batch:p1|genotype:Wild type TLAB,wt p1 S2,Bases 1 8 the UMI and template switch were trimmed from the remaining reads  trimmed stretches of polyA or polyT longer than 6bp and removed any reads that were shorter than 15 bp post trimming. We identified and depleted mtRNA and rRNA reads from the remaining reads by using Bowtie2 to align all reads against a transcriptome of mtRNA and rRNA  using the parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed   no unal. Remaining reads were then aligned to a modified version of the Ensembl Zv10 release 81 zebrafish transcriptome using Bowtie2 with parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed. The transcriptome start site of each gene was padded by 100bp to increase alignment rates  since reads were generated from the 5’ end of the transcript  and not all transcripts had correct annotations for their 5’ end. Additionally  pseudogenes were removed from the transcriptome as they sometimes interfered with real genes  given our 5’ coordinate selection described below. We chose the best alignment per read as the read with the most 5’ alignment coordinate to a transcript among those that aligned to the correct strand and had an alignment score of no more than 4 below the best alignment score overall. We then iteratively collapsed pairs of UMIs  if the sum of the minimum Phred quality scores across non matching bases was at most 30 to account for mis matched bases that were likely due to sequencing errors. A collapsed pair was represented for future comparisons by the UMI with a larger prior number of aligned reads unless the other UMI had at least 75% as many reads aligned and had equal or better UMI quality scores for those alignments evaluated as the softmin over aligned reads of the softmax over quality scores for the UMI bases in an aligned read. post collapsing  the number of transcripts observed per gene was calculated. This value was then adjusted based on the probability of a collision the drawing at random of the same UMI multiple times  which was calculated from the number of UMIs not detected for a given gene. The full length fragmented reads without xxx’ enrichment or UMI were aligned using RSEM against a reference transcriptome  which is depleted of mtRNA  rRNA  and pseudogenes as described above. The command ‘rsem calculate expression’ was used with parameters  p 4   paired end   seed length 24 for performing the alignment. The resulting gene levels in FPKM was used to build the expression matrix. Genome build: GRCz10 dr81 Supplementary files format and content: Tab delimited text files include gene levels UMI counts or FKPM for each sample.,Embryonic cell,,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for 5’ enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,Fish of each genotype were incrossed and embryos were collected. Embryos were dechorionated and allowed to grow at 28°C in standard fish water containing methylene blue till 50% epiboly stage.,developmental stage:50% epiboly|library preparation batch:p1|genotype:Wild type TLAB,GSM2838410,GSM2838410: wt p1 S2; Danio rerio; RNA Seq,GSM2838410,,1,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for five prime enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina HiSeq 2500,,SRP123526,,,wt_p1_S2.R1.fastq.gz wt_p1_S2.R2.fastq.gz,fastq fastq,88254208.0,1575968.0,GSM2838410 r1,0:24 1:32,A:24065563;C:19830888;G:20138155;T:24147163;N:72439,24,32,,,24065563,19830888,20138155,24147163,72439,SRX3357247,SRS2656328,SRA627905,GEO,"Schier Lab, Molecular and Cellular Biology, Harvard University",2,0.76816,0.81423,0.11267,0.11725,0.83956,0.83913,0.72098,0.72128,24,32,B,B,biological fallback assumption,illumina,hiseq_era,full_length,poly_a,unknown,sc,single_cell_plate,smartseq,,United States,2017-11-02,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
44127,SRR6250323,SRX3357246,SRS2656327,SRP123526,PRJNA416939,Single cell RNAseq SMART seq2 of wild type TLAB and MZoep tz57 zebrafish embryos at 50% epiboly stage,GSE106466,Transcriptome Analysis,SMART seq2 was performed on single cells isolated from visually staged zebrafish embryos. Overall design: Samples were all sequenced in one batch. Some were generated with a five prime' UMI tagged method  and others are full length SMART seq2.,parent bioproject:PRJNA417291,pubmed:29700225,,wt p1 S1,GSM2838409,,tissue:Embryonic cell|developmental stage:50% epiboly|library preparation batch:p1|genotype:Wild type TLAB,wt p1 S1,Bases 1 8 the UMI and template switch were trimmed from the remaining reads  trimmed stretches of polyA or polyT longer than 6bp and removed any reads that were shorter than 15 bp post trimming. We identified and depleted mtRNA and rRNA reads from the remaining reads by using Bowtie2 to align all reads against a transcriptome of mtRNA and rRNA  using the parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed   no unal. Remaining reads were then aligned to a modified version of the Ensembl Zv10 release 81 zebrafish transcriptome using Bowtie2 with parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed. The transcriptome start site of each gene was padded by 100bp to increase alignment rates  since reads were generated from the 5’ end of the transcript  and not all transcripts had correct annotations for their 5’ end. Additionally  pseudogenes were removed from the transcriptome as they sometimes interfered with real genes  given our 5’ coordinate selection described below. We chose the best alignment per read as the read with the most 5’ alignment coordinate to a transcript among those that aligned to the correct strand and had an alignment score of no more than 4 below the best alignment score overall. We then iteratively collapsed pairs of UMIs  if the sum of the minimum Phred quality scores across non matching bases was at most 30 to account for mis matched bases that were likely due to sequencing errors. A collapsed pair was represented for future comparisons by the UMI with a larger prior number of aligned reads unless the other UMI had at least 75% as many reads aligned and had equal or better UMI quality scores for those alignments evaluated as the softmin over aligned reads of the softmax over quality scores for the UMI bases in an aligned read. post collapsing  the number of transcripts observed per gene was calculated. This value was then adjusted based on the probability of a collision the drawing at random of the same UMI multiple times  which was calculated from the number of UMIs not detected for a given gene. The full length fragmented reads without xxx’ enrichment or UMI were aligned using RSEM against a reference transcriptome  which is depleted of mtRNA  rRNA  and pseudogenes as described above. The command ‘rsem calculate expression’ was used with parameters  p 4   paired end   seed length 24 for performing the alignment. The resulting gene levels in FPKM was used to build the expression matrix. Genome build: GRCz10 dr81 Supplementary files format and content: Tab delimited text files include gene levels UMI counts or FKPM for each sample.,Embryonic cell,,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for 5’ enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,Fish of each genotype were incrossed and embryos were collected. Embryos were dechorionated and allowed to grow at 28°C in standard fish water containing methylene blue till 50% epiboly stage.,developmental stage:50% epiboly|library preparation batch:p1|genotype:Wild type TLAB,GSM2838409,GSM2838409: wt p1 S1; Danio rerio; RNA Seq,GSM2838409,,1,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for five prime enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina HiSeq 2500,,SRP123526,,,wt_p1_S1.R1.fastq.gz wt_p1_S1.R2.fastq.gz,fastq fastq,69214768.0,1235978.0,GSM2838409 r1,0:24 1:32,A:19113770;C:15404084;G:15514812;T:19124524;N:57578,24,32,,,19113770,15404084,15514812,19124524,57578,SRX3357246,SRS2656327,SRA627905,GEO,"Schier Lab, Molecular and Cellular Biology, Harvard University",2,0.78022,0.83006,0.09266,0.09517,0.87032,0.87056,0.74712,0.74427,24,32,B,B,biological fallback assumption,illumina,hiseq_era,full_length,poly_a,unknown,sc,single_cell_plate,smartseq,,United States,2017-11-02,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
44128,SRR6250322,SRX3357245,SRS2656326,SRP123526,PRJNA416939,Single cell RNAseq SMART seq2 of wild type TLAB and MZoep tz57 zebrafish embryos at 50% epiboly stage,GSE106466,Transcriptome Analysis,SMART seq2 was performed on single cells isolated from visually staged zebrafish embryos. Overall design: Samples were all sequenced in one batch. Some were generated with a five prime' UMI tagged method  and others are full length SMART seq2.,parent bioproject:PRJNA417291,pubmed:29700225,,oep p4 S9,GSM2838408,,tissue:Embryonic cell|developmental stage:50% epiboly|library preparation batch:p4|genotype:MZoep tz57,oep p4 S9,Bases 1 8 the UMI and template switch were trimmed from the remaining reads  trimmed stretches of polyA or polyT longer than 6bp and removed any reads that were shorter than 15 bp post trimming. We identified and depleted mtRNA and rRNA reads from the remaining reads by using Bowtie2 to align all reads against a transcriptome of mtRNA and rRNA  using the parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed   no unal. Remaining reads were then aligned to a modified version of the Ensembl Zv10 release 81 zebrafish transcriptome using Bowtie2 with parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed. The transcriptome start site of each gene was padded by 100bp to increase alignment rates  since reads were generated from the 5’ end of the transcript  and not all transcripts had correct annotations for their 5’ end. Additionally  pseudogenes were removed from the transcriptome as they sometimes interfered with real genes  given our 5’ coordinate selection described below. We chose the best alignment per read as the read with the most 5’ alignment coordinate to a transcript among those that aligned to the correct strand and had an alignment score of no more than 4 below the best alignment score overall. We then iteratively collapsed pairs of UMIs  if the sum of the minimum Phred quality scores across non matching bases was at most 30 to account for mis matched bases that were likely due to sequencing errors. A collapsed pair was represented for future comparisons by the UMI with a larger prior number of aligned reads unless the other UMI had at least 75% as many reads aligned and had equal or better UMI quality scores for those alignments evaluated as the softmin over aligned reads of the softmax over quality scores for the UMI bases in an aligned read. post collapsing  the number of transcripts observed per gene was calculated. This value was then adjusted based on the probability of a collision the drawing at random of the same UMI multiple times  which was calculated from the number of UMIs not detected for a given gene. The full length fragmented reads without xxx’ enrichment or UMI were aligned using RSEM against a reference transcriptome  which is depleted of mtRNA  rRNA  and pseudogenes as described above. The command ‘rsem calculate expression’ was used with parameters  p 4   paired end   seed length 24 for performing the alignment. The resulting gene levels in FPKM was used to build the expression matrix. Genome build: GRCz10 dr81 Supplementary files format and content: Tab delimited text files include gene levels UMI counts or FKPM for each sample.,Embryonic cell,,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for 5’ enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,Fish of each genotype were incrossed and embryos were collected. Embryos were dechorionated and allowed to grow at 28°C in standard fish water containing methylene blue till 50% epiboly stage.,developmental stage:50% epiboly|library preparation batch:p4|genotype:MZoep tz57,GSM2838408,GSM2838408: oep p4 S9; Danio rerio; RNA Seq,GSM2838408,,1,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for five prime enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina HiSeq 2500,,SRP123526,,,oep_p4_S9.R1.fastq.gz oep_p4_S9.R2.fastq.gz,fastq fastq,59607240.0,1064415.0,GSM2838408 r1,0:24 1:32,A:15889806;C:13717017;G:13980126;T:15971426;N:48865,24,32,,,15889806,13717017,13980126,15971426,48865,SRX3357245,SRS2656326,SRA627905,GEO,"Schier Lab, Molecular and Cellular Biology, Harvard University",2,0.70468,0.74568,0.12923,0.13658,0.87969,0.87998,0.75196,0.7565,24,32,B,B,biological fallback assumption,illumina,hiseq_era,full_length,poly_a,unknown,sc,single_cell_plate,smartseq,,United States,2017-11-02,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
44129,SRR6250321,SRX3357244,SRS2656325,SRP123526,PRJNA416939,Single cell RNAseq SMART seq2 of wild type TLAB and MZoep tz57 zebrafish embryos at 50% epiboly stage,GSE106466,Transcriptome Analysis,SMART seq2 was performed on single cells isolated from visually staged zebrafish embryos. Overall design: Samples were all sequenced in one batch. Some were generated with a five prime' UMI tagged method  and others are full length SMART seq2.,parent bioproject:PRJNA417291,pubmed:29700225,,oep p4 S8,GSM2838407,,tissue:Embryonic cell|developmental stage:50% epiboly|library preparation batch:p4|genotype:MZoep tz57,oep p4 S8,Bases 1 8 the UMI and template switch were trimmed from the remaining reads  trimmed stretches of polyA or polyT longer than 6bp and removed any reads that were shorter than 15 bp post trimming. We identified and depleted mtRNA and rRNA reads from the remaining reads by using Bowtie2 to align all reads against a transcriptome of mtRNA and rRNA  using the parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed   no unal. Remaining reads were then aligned to a modified version of the Ensembl Zv10 release 81 zebrafish transcriptome using Bowtie2 with parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed. The transcriptome start site of each gene was padded by 100bp to increase alignment rates  since reads were generated from the 5’ end of the transcript  and not all transcripts had correct annotations for their 5’ end. Additionally  pseudogenes were removed from the transcriptome as they sometimes interfered with real genes  given our 5’ coordinate selection described below. We chose the best alignment per read as the read with the most 5’ alignment coordinate to a transcript among those that aligned to the correct strand and had an alignment score of no more than 4 below the best alignment score overall. We then iteratively collapsed pairs of UMIs  if the sum of the minimum Phred quality scores across non matching bases was at most 30 to account for mis matched bases that were likely due to sequencing errors. A collapsed pair was represented for future comparisons by the UMI with a larger prior number of aligned reads unless the other UMI had at least 75% as many reads aligned and had equal or better UMI quality scores for those alignments evaluated as the softmin over aligned reads of the softmax over quality scores for the UMI bases in an aligned read. post collapsing  the number of transcripts observed per gene was calculated. This value was then adjusted based on the probability of a collision the drawing at random of the same UMI multiple times  which was calculated from the number of UMIs not detected for a given gene. The full length fragmented reads without xxx’ enrichment or UMI were aligned using RSEM against a reference transcriptome  which is depleted of mtRNA  rRNA  and pseudogenes as described above. The command ‘rsem calculate expression’ was used with parameters  p 4   paired end   seed length 24 for performing the alignment. The resulting gene levels in FPKM was used to build the expression matrix. Genome build: GRCz10 dr81 Supplementary files format and content: Tab delimited text files include gene levels UMI counts or FKPM for each sample.,Embryonic cell,,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for 5’ enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,Fish of each genotype were incrossed and embryos were collected. Embryos were dechorionated and allowed to grow at 28°C in standard fish water containing methylene blue till 50% epiboly stage.,developmental stage:50% epiboly|library preparation batch:p4|genotype:MZoep tz57,GSM2838407,GSM2838407: oep p4 S8; Danio rerio; RNA Seq,GSM2838407,,1,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for five prime enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina HiSeq 2500,,SRP123526,,,oep_p4_S8.R1.fastq.gz oep_p4_S8.R2.fastq.gz,fastq fastq,90621608.0,1618243.0,GSM2838407 r1,0:24 1:32,A:23755286;C:21263521;G:21526665;T:23999805;N:76331,24,32,,,23755286,21263521,21526665,23999805,76331,SRX3357244,SRS2656325,SRA627905,GEO,"Schier Lab, Molecular and Cellular Biology, Harvard University",2,0.80318,0.85365,0.18029,0.19282,0.85283,0.85271,0.70396,0.69979,24,32,B,B,biological fallback assumption,illumina,hiseq_era,full_length,poly_a,unknown,sc,single_cell_plate,smartseq,,United States,2017-11-02,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
44130,SRR6250320,SRX3357243,SRS2656324,SRP123526,PRJNA416939,Single cell RNAseq SMART seq2 of wild type TLAB and MZoep tz57 zebrafish embryos at 50% epiboly stage,GSE106466,Transcriptome Analysis,SMART seq2 was performed on single cells isolated from visually staged zebrafish embryos. Overall design: Samples were all sequenced in one batch. Some were generated with a five prime' UMI tagged method  and others are full length SMART seq2.,parent bioproject:PRJNA417291,pubmed:29700225,,oep p4 S7,GSM2838406,,tissue:Embryonic cell|developmental stage:50% epiboly|library preparation batch:p4|genotype:MZoep tz57,oep p4 S7,Bases 1 8 the UMI and template switch were trimmed from the remaining reads  trimmed stretches of polyA or polyT longer than 6bp and removed any reads that were shorter than 15 bp post trimming. We identified and depleted mtRNA and rRNA reads from the remaining reads by using Bowtie2 to align all reads against a transcriptome of mtRNA and rRNA  using the parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed   no unal. Remaining reads were then aligned to a modified version of the Ensembl Zv10 release 81 zebrafish transcriptome using Bowtie2 with parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed. The transcriptome start site of each gene was padded by 100bp to increase alignment rates  since reads were generated from the 5’ end of the transcript  and not all transcripts had correct annotations for their 5’ end. Additionally  pseudogenes were removed from the transcriptome as they sometimes interfered with real genes  given our 5’ coordinate selection described below. We chose the best alignment per read as the read with the most 5’ alignment coordinate to a transcript among those that aligned to the correct strand and had an alignment score of no more than 4 below the best alignment score overall. We then iteratively collapsed pairs of UMIs  if the sum of the minimum Phred quality scores across non matching bases was at most 30 to account for mis matched bases that were likely due to sequencing errors. A collapsed pair was represented for future comparisons by the UMI with a larger prior number of aligned reads unless the other UMI had at least 75% as many reads aligned and had equal or better UMI quality scores for those alignments evaluated as the softmin over aligned reads of the softmax over quality scores for the UMI bases in an aligned read. post collapsing  the number of transcripts observed per gene was calculated. This value was then adjusted based on the probability of a collision the drawing at random of the same UMI multiple times  which was calculated from the number of UMIs not detected for a given gene. The full length fragmented reads without xxx’ enrichment or UMI were aligned using RSEM against a reference transcriptome  which is depleted of mtRNA  rRNA  and pseudogenes as described above. The command ‘rsem calculate expression’ was used with parameters  p 4   paired end   seed length 24 for performing the alignment. The resulting gene levels in FPKM was used to build the expression matrix. Genome build: GRCz10 dr81 Supplementary files format and content: Tab delimited text files include gene levels UMI counts or FKPM for each sample.,Embryonic cell,,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for 5’ enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,Fish of each genotype were incrossed and embryos were collected. Embryos were dechorionated and allowed to grow at 28°C in standard fish water containing methylene blue till 50% epiboly stage.,developmental stage:50% epiboly|library preparation batch:p4|genotype:MZoep tz57,GSM2838406,GSM2838406: oep p4 S7; Danio rerio; RNA Seq,GSM2838406,,1,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for five prime enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina HiSeq 2500,,SRP123526,,,oep_p4_S7.R1.fastq.gz oep_p4_S7.R2.fastq.gz,fastq fastq,4912656.0,87726.0,GSM2838406 r1,0:24 1:32,A:1332044;C:1105801;G:1114366;T:1356143;N:4302,24,32,,,1332044,1105801,1114366,1356143,4302,SRX3357243,SRS2656324,SRA627905,GEO,"Schier Lab, Molecular and Cellular Biology, Harvard University",2,0.79097,0.83957,0.1723,0.18179,0.89852,0.89674,0.75439,0.74413,24,32,B,B,biological fallback assumption,illumina,hiseq_era,full_length,poly_a,unknown,sc,single_cell_plate,smartseq,,United States,2017-11-02,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
44131,SRR6250319,SRX3357241,SRS2656323,SRP123526,PRJNA416939,Single cell RNAseq SMART seq2 of wild type TLAB and MZoep tz57 zebrafish embryos at 50% epiboly stage,GSE106466,Transcriptome Analysis,SMART seq2 was performed on single cells isolated from visually staged zebrafish embryos. Overall design: Samples were all sequenced in one batch. Some were generated with a five prime' UMI tagged method  and others are full length SMART seq2.,parent bioproject:PRJNA417291,pubmed:29700225,,oep p4 S6,GSM2838405,,tissue:Embryonic cell|developmental stage:50% epiboly|library preparation batch:p4|genotype:MZoep tz57,oep p4 S6,Bases 1 8 the UMI and template switch were trimmed from the remaining reads  trimmed stretches of polyA or polyT longer than 6bp and removed any reads that were shorter than 15 bp post trimming. We identified and depleted mtRNA and rRNA reads from the remaining reads by using Bowtie2 to align all reads against a transcriptome of mtRNA and rRNA  using the parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed   no unal. Remaining reads were then aligned to a modified version of the Ensembl Zv10 release 81 zebrafish transcriptome using Bowtie2 with parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed. The transcriptome start site of each gene was padded by 100bp to increase alignment rates  since reads were generated from the 5’ end of the transcript  and not all transcripts had correct annotations for their 5’ end. Additionally  pseudogenes were removed from the transcriptome as they sometimes interfered with real genes  given our 5’ coordinate selection described below. We chose the best alignment per read as the read with the most 5’ alignment coordinate to a transcript among those that aligned to the correct strand and had an alignment score of no more than 4 below the best alignment score overall. We then iteratively collapsed pairs of UMIs  if the sum of the minimum Phred quality scores across non matching bases was at most 30 to account for mis matched bases that were likely due to sequencing errors. A collapsed pair was represented for future comparisons by the UMI with a larger prior number of aligned reads unless the other UMI had at least 75% as many reads aligned and had equal or better UMI quality scores for those alignments evaluated as the softmin over aligned reads of the softmax over quality scores for the UMI bases in an aligned read. post collapsing  the number of transcripts observed per gene was calculated. This value was then adjusted based on the probability of a collision the drawing at random of the same UMI multiple times  which was calculated from the number of UMIs not detected for a given gene. The full length fragmented reads without xxx’ enrichment or UMI were aligned using RSEM against a reference transcriptome  which is depleted of mtRNA  rRNA  and pseudogenes as described above. The command ‘rsem calculate expression’ was used with parameters  p 4   paired end   seed length 24 for performing the alignment. The resulting gene levels in FPKM was used to build the expression matrix. Genome build: GRCz10 dr81 Supplementary files format and content: Tab delimited text files include gene levels UMI counts or FKPM for each sample.,Embryonic cell,,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for 5’ enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,Fish of each genotype were incrossed and embryos were collected. Embryos were dechorionated and allowed to grow at 28°C in standard fish water containing methylene blue till 50% epiboly stage.,developmental stage:50% epiboly|library preparation batch:p4|genotype:MZoep tz57,GSM2838405,GSM2838405: oep p4 S6; Danio rerio; RNA Seq,GSM2838405,,1,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for five prime enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina HiSeq 2500,,SRP123526,,,oep_p4_S6.R1.fastq.gz oep_p4_S6.R2.fastq.gz,fastq fastq,93200072.0,1664287.0,GSM2838405 r1,0:24 1:32,A:25640963;C:20843604;G:21078374;T:25559248;N:77883,24,32,,,25640963,20843604,21078374,25559248,77883,SRX3357241,SRS2656323,SRA627905,GEO,"Schier Lab, Molecular and Cellular Biology, Harvard University",2,0.80049,0.84734,0.10078,0.10503,0.88921,0.89045,0.76168,0.75227,24,32,B,B,biological fallback assumption,illumina,hiseq_era,full_length,poly_a,unknown,sc,single_cell_plate,smartseq,,United States,2017-11-02,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
44132,SRR6250318,SRX3357240,SRS2656321,SRP123526,PRJNA416939,Single cell RNAseq SMART seq2 of wild type TLAB and MZoep tz57 zebrafish embryos at 50% epiboly stage,GSE106466,Transcriptome Analysis,SMART seq2 was performed on single cells isolated from visually staged zebrafish embryos. Overall design: Samples were all sequenced in one batch. Some were generated with a five prime' UMI tagged method  and others are full length SMART seq2.,parent bioproject:PRJNA417291,pubmed:29700225,,oep p4 S5,GSM2838404,,tissue:Embryonic cell|developmental stage:50% epiboly|library preparation batch:p4|genotype:MZoep tz57,oep p4 S5,Bases 1 8 the UMI and template switch were trimmed from the remaining reads  trimmed stretches of polyA or polyT longer than 6bp and removed any reads that were shorter than 15 bp post trimming. We identified and depleted mtRNA and rRNA reads from the remaining reads by using Bowtie2 to align all reads against a transcriptome of mtRNA and rRNA  using the parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed   no unal. Remaining reads were then aligned to a modified version of the Ensembl Zv10 release 81 zebrafish transcriptome using Bowtie2 with parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed. The transcriptome start site of each gene was padded by 100bp to increase alignment rates  since reads were generated from the 5’ end of the transcript  and not all transcripts had correct annotations for their 5’ end. Additionally  pseudogenes were removed from the transcriptome as they sometimes interfered with real genes  given our 5’ coordinate selection described below. We chose the best alignment per read as the read with the most 5’ alignment coordinate to a transcript among those that aligned to the correct strand and had an alignment score of no more than 4 below the best alignment score overall. We then iteratively collapsed pairs of UMIs  if the sum of the minimum Phred quality scores across non matching bases was at most 30 to account for mis matched bases that were likely due to sequencing errors. A collapsed pair was represented for future comparisons by the UMI with a larger prior number of aligned reads unless the other UMI had at least 75% as many reads aligned and had equal or better UMI quality scores for those alignments evaluated as the softmin over aligned reads of the softmax over quality scores for the UMI bases in an aligned read. post collapsing  the number of transcripts observed per gene was calculated. This value was then adjusted based on the probability of a collision the drawing at random of the same UMI multiple times  which was calculated from the number of UMIs not detected for a given gene. The full length fragmented reads without xxx’ enrichment or UMI were aligned using RSEM against a reference transcriptome  which is depleted of mtRNA  rRNA  and pseudogenes as described above. The command ‘rsem calculate expression’ was used with parameters  p 4   paired end   seed length 24 for performing the alignment. The resulting gene levels in FPKM was used to build the expression matrix. Genome build: GRCz10 dr81 Supplementary files format and content: Tab delimited text files include gene levels UMI counts or FKPM for each sample.,Embryonic cell,,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for 5’ enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,Fish of each genotype were incrossed and embryos were collected. Embryos were dechorionated and allowed to grow at 28°C in standard fish water containing methylene blue till 50% epiboly stage.,developmental stage:50% epiboly|library preparation batch:p4|genotype:MZoep tz57,GSM2838404,GSM2838404: oep p4 S5; Danio rerio; RNA Seq,GSM2838404,,1,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for five prime enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina HiSeq 2500,,SRP123526,,,oep_p4_S5.R1.fastq.gz oep_p4_S5.R2.fastq.gz,fastq fastq,98958440.0,1767115.0,GSM2838404 r1,0:24 1:32,A:26086584;C:23066643;G:23403459;T:26320507;N:81247,24,32,,,26086584,23066643,23403459,26320507,81247,SRX3357240,SRS2656321,SRA627905,GEO,"Schier Lab, Molecular and Cellular Biology, Harvard University",2,0.79971,0.85115,0.18891,0.20331,0.85896,0.85995,0.71987,0.71776,24,32,B,B,biological fallback assumption,illumina,hiseq_era,full_length,poly_a,unknown,sc,single_cell_plate,smartseq,,United States,2017-11-02,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
44133,SRR6250317,SRX3357239,SRS2656319,SRP123526,PRJNA416939,Single cell RNAseq SMART seq2 of wild type TLAB and MZoep tz57 zebrafish embryos at 50% epiboly stage,GSE106466,Transcriptome Analysis,SMART seq2 was performed on single cells isolated from visually staged zebrafish embryos. Overall design: Samples were all sequenced in one batch. Some were generated with a five prime' UMI tagged method  and others are full length SMART seq2.,parent bioproject:PRJNA417291,pubmed:29700225,,oep p4 S4,GSM2838403,,tissue:Embryonic cell|developmental stage:50% epiboly|library preparation batch:p4|genotype:MZoep tz57,oep p4 S4,Bases 1 8 the UMI and template switch were trimmed from the remaining reads  trimmed stretches of polyA or polyT longer than 6bp and removed any reads that were shorter than 15 bp post trimming. We identified and depleted mtRNA and rRNA reads from the remaining reads by using Bowtie2 to align all reads against a transcriptome of mtRNA and rRNA  using the parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed   no unal. Remaining reads were then aligned to a modified version of the Ensembl Zv10 release 81 zebrafish transcriptome using Bowtie2 with parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed. The transcriptome start site of each gene was padded by 100bp to increase alignment rates  since reads were generated from the 5’ end of the transcript  and not all transcripts had correct annotations for their 5’ end. Additionally  pseudogenes were removed from the transcriptome as they sometimes interfered with real genes  given our 5’ coordinate selection described below. We chose the best alignment per read as the read with the most 5’ alignment coordinate to a transcript among those that aligned to the correct strand and had an alignment score of no more than 4 below the best alignment score overall. We then iteratively collapsed pairs of UMIs  if the sum of the minimum Phred quality scores across non matching bases was at most 30 to account for mis matched bases that were likely due to sequencing errors. A collapsed pair was represented for future comparisons by the UMI with a larger prior number of aligned reads unless the other UMI had at least 75% as many reads aligned and had equal or better UMI quality scores for those alignments evaluated as the softmin over aligned reads of the softmax over quality scores for the UMI bases in an aligned read. post collapsing  the number of transcripts observed per gene was calculated. This value was then adjusted based on the probability of a collision the drawing at random of the same UMI multiple times  which was calculated from the number of UMIs not detected for a given gene. The full length fragmented reads without xxx’ enrichment or UMI were aligned using RSEM against a reference transcriptome  which is depleted of mtRNA  rRNA  and pseudogenes as described above. The command ‘rsem calculate expression’ was used with parameters  p 4   paired end   seed length 24 for performing the alignment. The resulting gene levels in FPKM was used to build the expression matrix. Genome build: GRCz10 dr81 Supplementary files format and content: Tab delimited text files include gene levels UMI counts or FKPM for each sample.,Embryonic cell,,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for 5’ enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,Fish of each genotype were incrossed and embryos were collected. Embryos were dechorionated and allowed to grow at 28°C in standard fish water containing methylene blue till 50% epiboly stage.,developmental stage:50% epiboly|library preparation batch:p4|genotype:MZoep tz57,GSM2838403,GSM2838403: oep p4 S4; Danio rerio; RNA Seq,GSM2838403,,1,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for five prime enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina HiSeq 2500,,SRP123526,,,oep_p4_S4.R1.fastq.gz oep_p4_S4.R2.fastq.gz,fastq fastq,100896152.0,1801717.0,GSM2838403 r1,0:24 1:32,A:27173990;C:23092381;G:23338548;T:27207074;N:84159,24,32,,,27173990,23092381,23338548,27207074,84159,SRX3357239,SRS2656319,SRA627905,GEO,"Schier Lab, Molecular and Cellular Biology, Harvard University",2,0.80448,0.85347,0.12574,0.13447,0.86975,0.87125,0.76815,0.76891,24,32,B,B,biological fallback assumption,illumina,hiseq_era,full_length,poly_a,unknown,sc,single_cell_plate,smartseq,,United States,2017-11-02,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
44134,SRR6250316,SRX3357238,SRS2656320,SRP123526,PRJNA416939,Single cell RNAseq SMART seq2 of wild type TLAB and MZoep tz57 zebrafish embryos at 50% epiboly stage,GSE106466,Transcriptome Analysis,SMART seq2 was performed on single cells isolated from visually staged zebrafish embryos. Overall design: Samples were all sequenced in one batch. Some were generated with a five prime' UMI tagged method  and others are full length SMART seq2.,parent bioproject:PRJNA417291,pubmed:29700225,,oep p4 S3,GSM2838402,,tissue:Embryonic cell|developmental stage:50% epiboly|library preparation batch:p4|genotype:MZoep tz57,oep p4 S3,Bases 1 8 the UMI and template switch were trimmed from the remaining reads  trimmed stretches of polyA or polyT longer than 6bp and removed any reads that were shorter than 15 bp post trimming. We identified and depleted mtRNA and rRNA reads from the remaining reads by using Bowtie2 to align all reads against a transcriptome of mtRNA and rRNA  using the parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed   no unal. Remaining reads were then aligned to a modified version of the Ensembl Zv10 release 81 zebrafish transcriptome using Bowtie2 with parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed. The transcriptome start site of each gene was padded by 100bp to increase alignment rates  since reads were generated from the 5’ end of the transcript  and not all transcripts had correct annotations for their 5’ end. Additionally  pseudogenes were removed from the transcriptome as they sometimes interfered with real genes  given our 5’ coordinate selection described below. We chose the best alignment per read as the read with the most 5’ alignment coordinate to a transcript among those that aligned to the correct strand and had an alignment score of no more than 4 below the best alignment score overall. We then iteratively collapsed pairs of UMIs  if the sum of the minimum Phred quality scores across non matching bases was at most 30 to account for mis matched bases that were likely due to sequencing errors. A collapsed pair was represented for future comparisons by the UMI with a larger prior number of aligned reads unless the other UMI had at least 75% as many reads aligned and had equal or better UMI quality scores for those alignments evaluated as the softmin over aligned reads of the softmax over quality scores for the UMI bases in an aligned read. post collapsing  the number of transcripts observed per gene was calculated. This value was then adjusted based on the probability of a collision the drawing at random of the same UMI multiple times  which was calculated from the number of UMIs not detected for a given gene. The full length fragmented reads without xxx’ enrichment or UMI were aligned using RSEM against a reference transcriptome  which is depleted of mtRNA  rRNA  and pseudogenes as described above. The command ‘rsem calculate expression’ was used with parameters  p 4   paired end   seed length 24 for performing the alignment. The resulting gene levels in FPKM was used to build the expression matrix. Genome build: GRCz10 dr81 Supplementary files format and content: Tab delimited text files include gene levels UMI counts or FKPM for each sample.,Embryonic cell,,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for 5’ enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,Fish of each genotype were incrossed and embryos were collected. Embryos were dechorionated and allowed to grow at 28°C in standard fish water containing methylene blue till 50% epiboly stage.,developmental stage:50% epiboly|library preparation batch:p4|genotype:MZoep tz57,GSM2838402,GSM2838402: oep p4 S3; Danio rerio; RNA Seq,GSM2838402,,1,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for five prime enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina HiSeq 2500,,SRP123526,,,oep_p4_S3.R1.fastq.gz oep_p4_S3.R2.fastq.gz,fastq fastq,12183136.0,217556.0,GSM2838402 r1,0:24 1:32,A:3345072;C:2686784;G:2709001;T:3432792;N:9487,24,32,,,3345072,2686784,2709001,3432792,9487,SRX3357238,SRS2656320,SRA627905,GEO,"Schier Lab, Molecular and Cellular Biology, Harvard University",2,0.79372,0.84371,0.14527,0.15384,0.87071,0.86916,0.76333,0.76589,24,32,B,B,biological fallback assumption,illumina,hiseq_era,full_length,poly_a,unknown,sc,single_cell_plate,smartseq,,United States,2017-11-02,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
44135,SRR6250315,SRX3357237,SRS2656318,SRP123526,PRJNA416939,Single cell RNAseq SMART seq2 of wild type TLAB and MZoep tz57 zebrafish embryos at 50% epiboly stage,GSE106466,Transcriptome Analysis,SMART seq2 was performed on single cells isolated from visually staged zebrafish embryos. Overall design: Samples were all sequenced in one batch. Some were generated with a five prime' UMI tagged method  and others are full length SMART seq2.,parent bioproject:PRJNA417291,pubmed:29700225,,oep p4 S2,GSM2838401,,tissue:Embryonic cell|developmental stage:50% epiboly|library preparation batch:p4|genotype:MZoep tz57,oep p4 S2,Bases 1 8 the UMI and template switch were trimmed from the remaining reads  trimmed stretches of polyA or polyT longer than 6bp and removed any reads that were shorter than 15 bp post trimming. We identified and depleted mtRNA and rRNA reads from the remaining reads by using Bowtie2 to align all reads against a transcriptome of mtRNA and rRNA  using the parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed   no unal. Remaining reads were then aligned to a modified version of the Ensembl Zv10 release 81 zebrafish transcriptome using Bowtie2 with parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed. The transcriptome start site of each gene was padded by 100bp to increase alignment rates  since reads were generated from the 5’ end of the transcript  and not all transcripts had correct annotations for their 5’ end. Additionally  pseudogenes were removed from the transcriptome as they sometimes interfered with real genes  given our 5’ coordinate selection described below. We chose the best alignment per read as the read with the most 5’ alignment coordinate to a transcript among those that aligned to the correct strand and had an alignment score of no more than 4 below the best alignment score overall. We then iteratively collapsed pairs of UMIs  if the sum of the minimum Phred quality scores across non matching bases was at most 30 to account for mis matched bases that were likely due to sequencing errors. A collapsed pair was represented for future comparisons by the UMI with a larger prior number of aligned reads unless the other UMI had at least 75% as many reads aligned and had equal or better UMI quality scores for those alignments evaluated as the softmin over aligned reads of the softmax over quality scores for the UMI bases in an aligned read. post collapsing  the number of transcripts observed per gene was calculated. This value was then adjusted based on the probability of a collision the drawing at random of the same UMI multiple times  which was calculated from the number of UMIs not detected for a given gene. The full length fragmented reads without xxx’ enrichment or UMI were aligned using RSEM against a reference transcriptome  which is depleted of mtRNA  rRNA  and pseudogenes as described above. The command ‘rsem calculate expression’ was used with parameters  p 4   paired end   seed length 24 for performing the alignment. The resulting gene levels in FPKM was used to build the expression matrix. Genome build: GRCz10 dr81 Supplementary files format and content: Tab delimited text files include gene levels UMI counts or FKPM for each sample.,Embryonic cell,,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for 5’ enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,Fish of each genotype were incrossed and embryos were collected. Embryos were dechorionated and allowed to grow at 28°C in standard fish water containing methylene blue till 50% epiboly stage.,developmental stage:50% epiboly|library preparation batch:p4|genotype:MZoep tz57,GSM2838401,GSM2838401: oep p4 S2; Danio rerio; RNA Seq,GSM2838401,,1,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for five prime enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina HiSeq 2500,,SRP123526,,,oep_p4_S2.R1.fastq.gz oep_p4_S2.R2.fastq.gz,fastq fastq,107148720.0,1913370.0,GSM2838401 r1,0:24 1:32,A:28268122;C:25019759;G:25347800;T:28424142;N:88897,24,32,,,28268122,25019759,25347800,28424142,88897,SRX3357237,SRS2656318,SRA627905,GEO,"Schier Lab, Molecular and Cellular Biology, Harvard University",2,0.8032,0.85403,0.16972,0.18136,0.8548,0.85409,0.72899,0.72574,24,32,B,B,biological fallback assumption,illumina,hiseq_era,full_length,poly_a,unknown,sc,single_cell_plate,smartseq,,United States,2017-11-02,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
44136,SRR6250314,SRX3357236,SRS2656399,SRP123526,PRJNA416939,Single cell RNAseq SMART seq2 of wild type TLAB and MZoep tz57 zebrafish embryos at 50% epiboly stage,GSE106466,Transcriptome Analysis,SMART seq2 was performed on single cells isolated from visually staged zebrafish embryos. Overall design: Samples were all sequenced in one batch. Some were generated with a five prime' UMI tagged method  and others are full length SMART seq2.,parent bioproject:PRJNA417291,pubmed:29700225,,oep p4 S27,GSM2838400,,tissue:Embryonic cell|developmental stage:50% epiboly|library preparation batch:p4|genotype:MZoep tz57,oep p4 S27,Bases 1 8 the UMI and template switch were trimmed from the remaining reads  trimmed stretches of polyA or polyT longer than 6bp and removed any reads that were shorter than 15 bp post trimming. We identified and depleted mtRNA and rRNA reads from the remaining reads by using Bowtie2 to align all reads against a transcriptome of mtRNA and rRNA  using the parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed   no unal. Remaining reads were then aligned to a modified version of the Ensembl Zv10 release 81 zebrafish transcriptome using Bowtie2 with parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed. The transcriptome start site of each gene was padded by 100bp to increase alignment rates  since reads were generated from the 5’ end of the transcript  and not all transcripts had correct annotations for their 5’ end. Additionally  pseudogenes were removed from the transcriptome as they sometimes interfered with real genes  given our 5’ coordinate selection described below. We chose the best alignment per read as the read with the most 5’ alignment coordinate to a transcript among those that aligned to the correct strand and had an alignment score of no more than 4 below the best alignment score overall. We then iteratively collapsed pairs of UMIs  if the sum of the minimum Phred quality scores across non matching bases was at most 30 to account for mis matched bases that were likely due to sequencing errors. A collapsed pair was represented for future comparisons by the UMI with a larger prior number of aligned reads unless the other UMI had at least 75% as many reads aligned and had equal or better UMI quality scores for those alignments evaluated as the softmin over aligned reads of the softmax over quality scores for the UMI bases in an aligned read. post collapsing  the number of transcripts observed per gene was calculated. This value was then adjusted based on the probability of a collision the drawing at random of the same UMI multiple times  which was calculated from the number of UMIs not detected for a given gene. The full length fragmented reads without xxx’ enrichment or UMI were aligned using RSEM against a reference transcriptome  which is depleted of mtRNA  rRNA  and pseudogenes as described above. The command ‘rsem calculate expression’ was used with parameters  p 4   paired end   seed length 24 for performing the alignment. The resulting gene levels in FPKM was used to build the expression matrix. Genome build: GRCz10 dr81 Supplementary files format and content: Tab delimited text files include gene levels UMI counts or FKPM for each sample.,Embryonic cell,,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for 5’ enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,Fish of each genotype were incrossed and embryos were collected. Embryos were dechorionated and allowed to grow at 28°C in standard fish water containing methylene blue till 50% epiboly stage.,developmental stage:50% epiboly|library preparation batch:p4|genotype:MZoep tz57,GSM2838400,GSM2838400: oep p4 S27; Danio rerio; RNA Seq,GSM2838400,,1,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for five prime enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina HiSeq 2500,,SRP123526,,,oep_p4_S27.R1.fastq.gz oep_p4_S27.R2.fastq.gz,fastq fastq,2688.0,48.0,GSM2838400 r1,0:24 1:32,A:651;C:681;G:654;T:671;N:31,24,32,,,651,681,654,671,31,SRX3357236,SRS2656399,SRA627905,GEO,"Schier Lab, Molecular and Cellular Biology, Harvard University",2,0.69231,0.72728,0.15384,0.36363,0.99965,0.99991,0.80952,0.75,24,32,B,B,biological fallback assumption,illumina,hiseq_era,full_length,poly_a,unknown,sc,single_cell_plate,smartseq,,United States,2017-11-02,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
44137,SRR6250313,SRX3357235,SRS2656317,SRP123526,PRJNA416939,Single cell RNAseq SMART seq2 of wild type TLAB and MZoep tz57 zebrafish embryos at 50% epiboly stage,GSE106466,Transcriptome Analysis,SMART seq2 was performed on single cells isolated from visually staged zebrafish embryos. Overall design: Samples were all sequenced in one batch. Some were generated with a five prime' UMI tagged method  and others are full length SMART seq2.,parent bioproject:PRJNA417291,pubmed:29700225,,oep p4 S26,GSM2838399,,tissue:Embryonic cell|developmental stage:50% epiboly|library preparation batch:p4|genotype:MZoep tz57,oep p4 S26,Bases 1 8 the UMI and template switch were trimmed from the remaining reads  trimmed stretches of polyA or polyT longer than 6bp and removed any reads that were shorter than 15 bp post trimming. We identified and depleted mtRNA and rRNA reads from the remaining reads by using Bowtie2 to align all reads against a transcriptome of mtRNA and rRNA  using the parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed   no unal. Remaining reads were then aligned to a modified version of the Ensembl Zv10 release 81 zebrafish transcriptome using Bowtie2 with parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed. The transcriptome start site of each gene was padded by 100bp to increase alignment rates  since reads were generated from the 5’ end of the transcript  and not all transcripts had correct annotations for their 5’ end. Additionally  pseudogenes were removed from the transcriptome as they sometimes interfered with real genes  given our 5’ coordinate selection described below. We chose the best alignment per read as the read with the most 5’ alignment coordinate to a transcript among those that aligned to the correct strand and had an alignment score of no more than 4 below the best alignment score overall. We then iteratively collapsed pairs of UMIs  if the sum of the minimum Phred quality scores across non matching bases was at most 30 to account for mis matched bases that were likely due to sequencing errors. A collapsed pair was represented for future comparisons by the UMI with a larger prior number of aligned reads unless the other UMI had at least 75% as many reads aligned and had equal or better UMI quality scores for those alignments evaluated as the softmin over aligned reads of the softmax over quality scores for the UMI bases in an aligned read. post collapsing  the number of transcripts observed per gene was calculated. This value was then adjusted based on the probability of a collision the drawing at random of the same UMI multiple times  which was calculated from the number of UMIs not detected for a given gene. The full length fragmented reads without xxx’ enrichment or UMI were aligned using RSEM against a reference transcriptome  which is depleted of mtRNA  rRNA  and pseudogenes as described above. The command ‘rsem calculate expression’ was used with parameters  p 4   paired end   seed length 24 for performing the alignment. The resulting gene levels in FPKM was used to build the expression matrix. Genome build: GRCz10 dr81 Supplementary files format and content: Tab delimited text files include gene levels UMI counts or FKPM for each sample.,Embryonic cell,,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for 5’ enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,Fish of each genotype were incrossed and embryos were collected. Embryos were dechorionated and allowed to grow at 28°C in standard fish water containing methylene blue till 50% epiboly stage.,developmental stage:50% epiboly|library preparation batch:p4|genotype:MZoep tz57,GSM2838399,GSM2838399: oep p4 S26; Danio rerio; RNA Seq,GSM2838399,,1,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for five prime enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina HiSeq 2500,,SRP123526,,,oep_p4_S26.R1.fastq.gz oep_p4_S26.R2.fastq.gz,fastq fastq,22568.0,403.0,GSM2838399 r1,0:24 1:32,A:6226;C:6218;G:5042;T:5082;N:0,24,32,,,6226,6218,5042,5082,0,SRX3357235,SRS2656317,SRA627905,GEO,"Schier Lab, Molecular and Cellular Biology, Harvard University",2,0.10363,0.21244,0.06994,0.13989,0.99979,0.99977,0.83333,0.84615,24,32,B,B,mate1-mate2 similar by mapping diff,illumina,hiseq_era,full_length,poly_a,unknown,sc,single_cell_plate,smartseq,,United States,2017-11-02,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
44138,SRR6250312,SRX3357234,SRS2656316,SRP123526,PRJNA416939,Single cell RNAseq SMART seq2 of wild type TLAB and MZoep tz57 zebrafish embryos at 50% epiboly stage,GSE106466,Transcriptome Analysis,SMART seq2 was performed on single cells isolated from visually staged zebrafish embryos. Overall design: Samples were all sequenced in one batch. Some were generated with a five prime' UMI tagged method  and others are full length SMART seq2.,parent bioproject:PRJNA417291,pubmed:29700225,,oep p4 S25,GSM2838398,,tissue:Embryonic cell|developmental stage:50% epiboly|library preparation batch:p4|genotype:MZoep tz57,oep p4 S25,Bases 1 8 the UMI and template switch were trimmed from the remaining reads  trimmed stretches of polyA or polyT longer than 6bp and removed any reads that were shorter than 15 bp post trimming. We identified and depleted mtRNA and rRNA reads from the remaining reads by using Bowtie2 to align all reads against a transcriptome of mtRNA and rRNA  using the parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed   no unal. Remaining reads were then aligned to a modified version of the Ensembl Zv10 release 81 zebrafish transcriptome using Bowtie2 with parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed. The transcriptome start site of each gene was padded by 100bp to increase alignment rates  since reads were generated from the 5’ end of the transcript  and not all transcripts had correct annotations for their 5’ end. Additionally  pseudogenes were removed from the transcriptome as they sometimes interfered with real genes  given our 5’ coordinate selection described below. We chose the best alignment per read as the read with the most 5’ alignment coordinate to a transcript among those that aligned to the correct strand and had an alignment score of no more than 4 below the best alignment score overall. We then iteratively collapsed pairs of UMIs  if the sum of the minimum Phred quality scores across non matching bases was at most 30 to account for mis matched bases that were likely due to sequencing errors. A collapsed pair was represented for future comparisons by the UMI with a larger prior number of aligned reads unless the other UMI had at least 75% as many reads aligned and had equal or better UMI quality scores for those alignments evaluated as the softmin over aligned reads of the softmax over quality scores for the UMI bases in an aligned read. post collapsing  the number of transcripts observed per gene was calculated. This value was then adjusted based on the probability of a collision the drawing at random of the same UMI multiple times  which was calculated from the number of UMIs not detected for a given gene. The full length fragmented reads without xxx’ enrichment or UMI were aligned using RSEM against a reference transcriptome  which is depleted of mtRNA  rRNA  and pseudogenes as described above. The command ‘rsem calculate expression’ was used with parameters  p 4   paired end   seed length 24 for performing the alignment. The resulting gene levels in FPKM was used to build the expression matrix. Genome build: GRCz10 dr81 Supplementary files format and content: Tab delimited text files include gene levels UMI counts or FKPM for each sample.,Embryonic cell,,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for 5’ enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,Fish of each genotype were incrossed and embryos were collected. Embryos were dechorionated and allowed to grow at 28°C in standard fish water containing methylene blue till 50% epiboly stage.,developmental stage:50% epiboly|library preparation batch:p4|genotype:MZoep tz57,GSM2838398,GSM2838398: oep p4 S25; Danio rerio; RNA Seq,GSM2838398,,1,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for five prime enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina HiSeq 2500,,SRP123526,,,oep_p4_S25.R1.fastq.gz oep_p4_S25.R2.fastq.gz,fastq fastq,54054560.0,965260.0,GSM2838398 r1,0:24 1:32,A:15083196;C:11824083;G:12012704;T:15089704;N:44873,24,32,,,15083196,11824083,12012704,15089704,44873,SRX3357234,SRS2656316,SRA627905,GEO,"Schier Lab, Molecular and Cellular Biology, Harvard University",2,0.77823,0.82959,0.09938,0.10266,0.87247,0.87261,0.75693,0.75981,24,32,B,B,biological fallback assumption,illumina,hiseq_era,full_length,poly_a,unknown,sc,single_cell_plate,smartseq,,United States,2017-11-02,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
44139,SRR6250311,SRX3357233,SRS2656315,SRP123526,PRJNA416939,Single cell RNAseq SMART seq2 of wild type TLAB and MZoep tz57 zebrafish embryos at 50% epiboly stage,GSE106466,Transcriptome Analysis,SMART seq2 was performed on single cells isolated from visually staged zebrafish embryos. Overall design: Samples were all sequenced in one batch. Some were generated with a five prime' UMI tagged method  and others are full length SMART seq2.,parent bioproject:PRJNA417291,pubmed:29700225,,oep p4 S24,GSM2838397,,tissue:Embryonic cell|developmental stage:50% epiboly|library preparation batch:p4|genotype:MZoep tz57,oep p4 S24,Bases 1 8 the UMI and template switch were trimmed from the remaining reads  trimmed stretches of polyA or polyT longer than 6bp and removed any reads that were shorter than 15 bp post trimming. We identified and depleted mtRNA and rRNA reads from the remaining reads by using Bowtie2 to align all reads against a transcriptome of mtRNA and rRNA  using the parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed   no unal. Remaining reads were then aligned to a modified version of the Ensembl Zv10 release 81 zebrafish transcriptome using Bowtie2 with parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed. The transcriptome start site of each gene was padded by 100bp to increase alignment rates  since reads were generated from the 5’ end of the transcript  and not all transcripts had correct annotations for their 5’ end. Additionally  pseudogenes were removed from the transcriptome as they sometimes interfered with real genes  given our 5’ coordinate selection described below. We chose the best alignment per read as the read with the most 5’ alignment coordinate to a transcript among those that aligned to the correct strand and had an alignment score of no more than 4 below the best alignment score overall. We then iteratively collapsed pairs of UMIs  if the sum of the minimum Phred quality scores across non matching bases was at most 30 to account for mis matched bases that were likely due to sequencing errors. A collapsed pair was represented for future comparisons by the UMI with a larger prior number of aligned reads unless the other UMI had at least 75% as many reads aligned and had equal or better UMI quality scores for those alignments evaluated as the softmin over aligned reads of the softmax over quality scores for the UMI bases in an aligned read. post collapsing  the number of transcripts observed per gene was calculated. This value was then adjusted based on the probability of a collision the drawing at random of the same UMI multiple times  which was calculated from the number of UMIs not detected for a given gene. The full length fragmented reads without xxx’ enrichment or UMI were aligned using RSEM against a reference transcriptome  which is depleted of mtRNA  rRNA  and pseudogenes as described above. The command ‘rsem calculate expression’ was used with parameters  p 4   paired end   seed length 24 for performing the alignment. The resulting gene levels in FPKM was used to build the expression matrix. Genome build: GRCz10 dr81 Supplementary files format and content: Tab delimited text files include gene levels UMI counts or FKPM for each sample.,Embryonic cell,,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for 5’ enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,Fish of each genotype were incrossed and embryos were collected. Embryos were dechorionated and allowed to grow at 28°C in standard fish water containing methylene blue till 50% epiboly stage.,developmental stage:50% epiboly|library preparation batch:p4|genotype:MZoep tz57,GSM2838397,GSM2838397: oep p4 S24; Danio rerio; RNA Seq,GSM2838397,,1,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for five prime enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina HiSeq 2500,,SRP123526,,,oep_p4_S24.R1.fastq.gz oep_p4_S24.R2.fastq.gz,fastq fastq,65370312.0,1167327.0,GSM2838397 r1,0:24 1:32,A:18166371;C:14411204;G:14686361;T:18054122;N:52254,24,32,,,18166371,14411204,14686361,18054122,52254,SRX3357233,SRS2656315,SRA627905,GEO,"Schier Lab, Molecular and Cellular Biology, Harvard University",2,0.77848,0.8339,0.06212,0.06417,0.88286,0.88258,0.77544,0.77309,24,32,B,B,biological fallback assumption,illumina,hiseq_era,full_length,poly_a,unknown,sc,single_cell_plate,smartseq,,United States,2017-11-02,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
44140,SRR6250310,SRX3357232,SRS2656314,SRP123526,PRJNA416939,Single cell RNAseq SMART seq2 of wild type TLAB and MZoep tz57 zebrafish embryos at 50% epiboly stage,GSE106466,Transcriptome Analysis,SMART seq2 was performed on single cells isolated from visually staged zebrafish embryos. Overall design: Samples were all sequenced in one batch. Some were generated with a five prime' UMI tagged method  and others are full length SMART seq2.,parent bioproject:PRJNA417291,pubmed:29700225,,oep p4 S23,GSM2838396,,tissue:Embryonic cell|developmental stage:50% epiboly|library preparation batch:p4|genotype:MZoep tz57,oep p4 S23,Bases 1 8 the UMI and template switch were trimmed from the remaining reads  trimmed stretches of polyA or polyT longer than 6bp and removed any reads that were shorter than 15 bp post trimming. We identified and depleted mtRNA and rRNA reads from the remaining reads by using Bowtie2 to align all reads against a transcriptome of mtRNA and rRNA  using the parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed   no unal. Remaining reads were then aligned to a modified version of the Ensembl Zv10 release 81 zebrafish transcriptome using Bowtie2 with parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed. The transcriptome start site of each gene was padded by 100bp to increase alignment rates  since reads were generated from the 5’ end of the transcript  and not all transcripts had correct annotations for their 5’ end. Additionally  pseudogenes were removed from the transcriptome as they sometimes interfered with real genes  given our 5’ coordinate selection described below. We chose the best alignment per read as the read with the most 5’ alignment coordinate to a transcript among those that aligned to the correct strand and had an alignment score of no more than 4 below the best alignment score overall. We then iteratively collapsed pairs of UMIs  if the sum of the minimum Phred quality scores across non matching bases was at most 30 to account for mis matched bases that were likely due to sequencing errors. A collapsed pair was represented for future comparisons by the UMI with a larger prior number of aligned reads unless the other UMI had at least 75% as many reads aligned and had equal or better UMI quality scores for those alignments evaluated as the softmin over aligned reads of the softmax over quality scores for the UMI bases in an aligned read. post collapsing  the number of transcripts observed per gene was calculated. This value was then adjusted based on the probability of a collision the drawing at random of the same UMI multiple times  which was calculated from the number of UMIs not detected for a given gene. The full length fragmented reads without xxx’ enrichment or UMI were aligned using RSEM against a reference transcriptome  which is depleted of mtRNA  rRNA  and pseudogenes as described above. The command ‘rsem calculate expression’ was used with parameters  p 4   paired end   seed length 24 for performing the alignment. The resulting gene levels in FPKM was used to build the expression matrix. Genome build: GRCz10 dr81 Supplementary files format and content: Tab delimited text files include gene levels UMI counts or FKPM for each sample.,Embryonic cell,,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for 5’ enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,Fish of each genotype were incrossed and embryos were collected. Embryos were dechorionated and allowed to grow at 28°C in standard fish water containing methylene blue till 50% epiboly stage.,developmental stage:50% epiboly|library preparation batch:p4|genotype:MZoep tz57,GSM2838396,GSM2838396: oep p4 S23; Danio rerio; RNA Seq,GSM2838396,,1,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for five prime enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina HiSeq 2500,,SRP123526,,,oep_p4_S23.R1.fastq.gz oep_p4_S23.R2.fastq.gz,fastq fastq,70100016.0,1251786.0,GSM2838396 r1,0:24 1:32,A:19312414;C:15561812;G:15849161;T:19318033;N:58596,24,32,,,19312414,15561812,15849161,19318033,58596,SRX3357232,SRS2656314,SRA627905,GEO,"Schier Lab, Molecular and Cellular Biology, Harvard University",2,0.7756,0.82639,0.09516,0.09729,0.85009,0.85107,0.70903,0.71057,24,32,B,B,biological fallback assumption,illumina,hiseq_era,full_length,poly_a,unknown,sc,single_cell_plate,smartseq,,United States,2017-11-02,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
44141,SRR6250309,SRX3357231,SRS2656312,SRP123526,PRJNA416939,Single cell RNAseq SMART seq2 of wild type TLAB and MZoep tz57 zebrafish embryos at 50% epiboly stage,GSE106466,Transcriptome Analysis,SMART seq2 was performed on single cells isolated from visually staged zebrafish embryos. Overall design: Samples were all sequenced in one batch. Some were generated with a five prime' UMI tagged method  and others are full length SMART seq2.,parent bioproject:PRJNA417291,pubmed:29700225,,oep p4 S22,GSM2838395,,tissue:Embryonic cell|developmental stage:50% epiboly|library preparation batch:p4|genotype:MZoep tz57,oep p4 S22,Bases 1 8 the UMI and template switch were trimmed from the remaining reads  trimmed stretches of polyA or polyT longer than 6bp and removed any reads that were shorter than 15 bp post trimming. We identified and depleted mtRNA and rRNA reads from the remaining reads by using Bowtie2 to align all reads against a transcriptome of mtRNA and rRNA  using the parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed   no unal. Remaining reads were then aligned to a modified version of the Ensembl Zv10 release 81 zebrafish transcriptome using Bowtie2 with parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed. The transcriptome start site of each gene was padded by 100bp to increase alignment rates  since reads were generated from the 5’ end of the transcript  and not all transcripts had correct annotations for their 5’ end. Additionally  pseudogenes were removed from the transcriptome as they sometimes interfered with real genes  given our 5’ coordinate selection described below. We chose the best alignment per read as the read with the most 5’ alignment coordinate to a transcript among those that aligned to the correct strand and had an alignment score of no more than 4 below the best alignment score overall. We then iteratively collapsed pairs of UMIs  if the sum of the minimum Phred quality scores across non matching bases was at most 30 to account for mis matched bases that were likely due to sequencing errors. A collapsed pair was represented for future comparisons by the UMI with a larger prior number of aligned reads unless the other UMI had at least 75% as many reads aligned and had equal or better UMI quality scores for those alignments evaluated as the softmin over aligned reads of the softmax over quality scores for the UMI bases in an aligned read. post collapsing  the number of transcripts observed per gene was calculated. This value was then adjusted based on the probability of a collision the drawing at random of the same UMI multiple times  which was calculated from the number of UMIs not detected for a given gene. The full length fragmented reads without xxx’ enrichment or UMI were aligned using RSEM against a reference transcriptome  which is depleted of mtRNA  rRNA  and pseudogenes as described above. The command ‘rsem calculate expression’ was used with parameters  p 4   paired end   seed length 24 for performing the alignment. The resulting gene levels in FPKM was used to build the expression matrix. Genome build: GRCz10 dr81 Supplementary files format and content: Tab delimited text files include gene levels UMI counts or FKPM for each sample.,Embryonic cell,,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for 5’ enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,Fish of each genotype were incrossed and embryos were collected. Embryos were dechorionated and allowed to grow at 28°C in standard fish water containing methylene blue till 50% epiboly stage.,developmental stage:50% epiboly|library preparation batch:p4|genotype:MZoep tz57,GSM2838395,GSM2838395: oep p4 S22; Danio rerio; RNA Seq,GSM2838395,,1,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for five prime enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina HiSeq 2500,,SRP123526,,,oep_p4_S22.R1.fastq.gz oep_p4_S22.R2.fastq.gz,fastq fastq,4747120.0,84770.0,GSM2838395 r1,0:24 1:32,A:1288548;C:1060321;G:1081126;T:1313197;N:3928,24,32,,,1288548,1060321,1081126,1313197,3928,SRX3357231,SRS2656312,SRA627905,GEO,"Schier Lab, Molecular and Cellular Biology, Harvard University",2,0.76295,0.81806,0.13024,0.13447,0.90376,0.90264,0.73271,0.73327,24,32,B,B,biological fallback assumption,illumina,hiseq_era,full_length,poly_a,unknown,sc,single_cell_plate,smartseq,,United States,2017-11-02,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
44142,SRR6250308,SRX3357230,SRS2656313,SRP123526,PRJNA416939,Single cell RNAseq SMART seq2 of wild type TLAB and MZoep tz57 zebrafish embryos at 50% epiboly stage,GSE106466,Transcriptome Analysis,SMART seq2 was performed on single cells isolated from visually staged zebrafish embryos. Overall design: Samples were all sequenced in one batch. Some were generated with a five prime' UMI tagged method  and others are full length SMART seq2.,parent bioproject:PRJNA417291,pubmed:29700225,,oep p4 S21,GSM2838394,,tissue:Embryonic cell|developmental stage:50% epiboly|library preparation batch:p4|genotype:MZoep tz57,oep p4 S21,Bases 1 8 the UMI and template switch were trimmed from the remaining reads  trimmed stretches of polyA or polyT longer than 6bp and removed any reads that were shorter than 15 bp post trimming. We identified and depleted mtRNA and rRNA reads from the remaining reads by using Bowtie2 to align all reads against a transcriptome of mtRNA and rRNA  using the parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed   no unal. Remaining reads were then aligned to a modified version of the Ensembl Zv10 release 81 zebrafish transcriptome using Bowtie2 with parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed. The transcriptome start site of each gene was padded by 100bp to increase alignment rates  since reads were generated from the 5’ end of the transcript  and not all transcripts had correct annotations for their 5’ end. Additionally  pseudogenes were removed from the transcriptome as they sometimes interfered with real genes  given our 5’ coordinate selection described below. We chose the best alignment per read as the read with the most 5’ alignment coordinate to a transcript among those that aligned to the correct strand and had an alignment score of no more than 4 below the best alignment score overall. We then iteratively collapsed pairs of UMIs  if the sum of the minimum Phred quality scores across non matching bases was at most 30 to account for mis matched bases that were likely due to sequencing errors. A collapsed pair was represented for future comparisons by the UMI with a larger prior number of aligned reads unless the other UMI had at least 75% as many reads aligned and had equal or better UMI quality scores for those alignments evaluated as the softmin over aligned reads of the softmax over quality scores for the UMI bases in an aligned read. post collapsing  the number of transcripts observed per gene was calculated. This value was then adjusted based on the probability of a collision the drawing at random of the same UMI multiple times  which was calculated from the number of UMIs not detected for a given gene. The full length fragmented reads without xxx’ enrichment or UMI were aligned using RSEM against a reference transcriptome  which is depleted of mtRNA  rRNA  and pseudogenes as described above. The command ‘rsem calculate expression’ was used with parameters  p 4   paired end   seed length 24 for performing the alignment. The resulting gene levels in FPKM was used to build the expression matrix. Genome build: GRCz10 dr81 Supplementary files format and content: Tab delimited text files include gene levels UMI counts or FKPM for each sample.,Embryonic cell,,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for 5’ enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,Fish of each genotype were incrossed and embryos were collected. Embryos were dechorionated and allowed to grow at 28°C in standard fish water containing methylene blue till 50% epiboly stage.,developmental stage:50% epiboly|library preparation batch:p4|genotype:MZoep tz57,GSM2838394,GSM2838394: oep p4 S21; Danio rerio; RNA Seq,GSM2838394,,1,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for five prime enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina HiSeq 2500,,SRP123526,,,oep_p4_S21.R1.fastq.gz oep_p4_S21.R2.fastq.gz,fastq fastq,69224624.0,1236154.0,GSM2838394 r1,0:24 1:32,A:19150005;C:15344900;G:15612847;T:19059784;N:57088,24,32,,,19150005,15344900,15612847,19059784,57088,SRX3357230,SRS2656313,SRA627905,GEO,"Schier Lab, Molecular and Cellular Biology, Harvard University",2,0.78186,0.83463,0.08972,0.09272,0.86523,0.8648,0.76414,0.7627,24,32,B,B,biological fallback assumption,illumina,hiseq_era,full_length,poly_a,unknown,sc,single_cell_plate,smartseq,,United States,2017-11-02,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
44143,SRR6250307,SRX3357229,SRS2656311,SRP123526,PRJNA416939,Single cell RNAseq SMART seq2 of wild type TLAB and MZoep tz57 zebrafish embryos at 50% epiboly stage,GSE106466,Transcriptome Analysis,SMART seq2 was performed on single cells isolated from visually staged zebrafish embryos. Overall design: Samples were all sequenced in one batch. Some were generated with a five prime' UMI tagged method  and others are full length SMART seq2.,parent bioproject:PRJNA417291,pubmed:29700225,,oep p4 S20,GSM2838393,,tissue:Embryonic cell|developmental stage:50% epiboly|library preparation batch:p4|genotype:MZoep tz57,oep p4 S20,Bases 1 8 the UMI and template switch were trimmed from the remaining reads  trimmed stretches of polyA or polyT longer than 6bp and removed any reads that were shorter than 15 bp post trimming. We identified and depleted mtRNA and rRNA reads from the remaining reads by using Bowtie2 to align all reads against a transcriptome of mtRNA and rRNA  using the parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed   no unal. Remaining reads were then aligned to a modified version of the Ensembl Zv10 release 81 zebrafish transcriptome using Bowtie2 with parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed. The transcriptome start site of each gene was padded by 100bp to increase alignment rates  since reads were generated from the 5’ end of the transcript  and not all transcripts had correct annotations for their 5’ end. Additionally  pseudogenes were removed from the transcriptome as they sometimes interfered with real genes  given our 5’ coordinate selection described below. We chose the best alignment per read as the read with the most 5’ alignment coordinate to a transcript among those that aligned to the correct strand and had an alignment score of no more than 4 below the best alignment score overall. We then iteratively collapsed pairs of UMIs  if the sum of the minimum Phred quality scores across non matching bases was at most 30 to account for mis matched bases that were likely due to sequencing errors. A collapsed pair was represented for future comparisons by the UMI with a larger prior number of aligned reads unless the other UMI had at least 75% as many reads aligned and had equal or better UMI quality scores for those alignments evaluated as the softmin over aligned reads of the softmax over quality scores for the UMI bases in an aligned read. post collapsing  the number of transcripts observed per gene was calculated. This value was then adjusted based on the probability of a collision the drawing at random of the same UMI multiple times  which was calculated from the number of UMIs not detected for a given gene. The full length fragmented reads without xxx’ enrichment or UMI were aligned using RSEM against a reference transcriptome  which is depleted of mtRNA  rRNA  and pseudogenes as described above. The command ‘rsem calculate expression’ was used with parameters  p 4   paired end   seed length 24 for performing the alignment. The resulting gene levels in FPKM was used to build the expression matrix. Genome build: GRCz10 dr81 Supplementary files format and content: Tab delimited text files include gene levels UMI counts or FKPM for each sample.,Embryonic cell,,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for 5’ enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,Fish of each genotype were incrossed and embryos were collected. Embryos were dechorionated and allowed to grow at 28°C in standard fish water containing methylene blue till 50% epiboly stage.,developmental stage:50% epiboly|library preparation batch:p4|genotype:MZoep tz57,GSM2838393,GSM2838393: oep p4 S20; Danio rerio; RNA Seq,GSM2838393,,1,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for five prime enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina HiSeq 2500,,SRP123526,,,oep_p4_S20.R1.fastq.gz oep_p4_S20.R2.fastq.gz,fastq fastq,63875952.0,1140642.0,GSM2838393 r1,0:24 1:32,A:17811278;C:14030742;G:14262763;T:17718107;N:53062,24,32,,,17811278,14030742,14262763,17718107,53062,SRX3357229,SRS2656311,SRA627905,GEO,"Schier Lab, Molecular and Cellular Biology, Harvard University",2,0.782,0.83525,0.06874,0.07014,0.86618,0.86671,0.77992,0.78773,24,32,B,B,biological fallback assumption,illumina,hiseq_era,full_length,poly_a,unknown,sc,single_cell_plate,smartseq,,United States,2017-11-02,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
44144,SRR6250306,SRX3357228,SRS2656310,SRP123526,PRJNA416939,Single cell RNAseq SMART seq2 of wild type TLAB and MZoep tz57 zebrafish embryos at 50% epiboly stage,GSE106466,Transcriptome Analysis,SMART seq2 was performed on single cells isolated from visually staged zebrafish embryos. Overall design: Samples were all sequenced in one batch. Some were generated with a five prime' UMI tagged method  and others are full length SMART seq2.,parent bioproject:PRJNA417291,pubmed:29700225,,oep p4 S1,GSM2838392,,tissue:Embryonic cell|developmental stage:50% epiboly|library preparation batch:p4|genotype:MZoep tz57,oep p4 S1,Bases 1 8 the UMI and template switch were trimmed from the remaining reads  trimmed stretches of polyA or polyT longer than 6bp and removed any reads that were shorter than 15 bp post trimming. We identified and depleted mtRNA and rRNA reads from the remaining reads by using Bowtie2 to align all reads against a transcriptome of mtRNA and rRNA  using the parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed   no unal. Remaining reads were then aligned to a modified version of the Ensembl Zv10 release 81 zebrafish transcriptome using Bowtie2 with parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed. The transcriptome start site of each gene was padded by 100bp to increase alignment rates  since reads were generated from the 5’ end of the transcript  and not all transcripts had correct annotations for their 5’ end. Additionally  pseudogenes were removed from the transcriptome as they sometimes interfered with real genes  given our 5’ coordinate selection described below. We chose the best alignment per read as the read with the most 5’ alignment coordinate to a transcript among those that aligned to the correct strand and had an alignment score of no more than 4 below the best alignment score overall. We then iteratively collapsed pairs of UMIs  if the sum of the minimum Phred quality scores across non matching bases was at most 30 to account for mis matched bases that were likely due to sequencing errors. A collapsed pair was represented for future comparisons by the UMI with a larger prior number of aligned reads unless the other UMI had at least 75% as many reads aligned and had equal or better UMI quality scores for those alignments evaluated as the softmin over aligned reads of the softmax over quality scores for the UMI bases in an aligned read. post collapsing  the number of transcripts observed per gene was calculated. This value was then adjusted based on the probability of a collision the drawing at random of the same UMI multiple times  which was calculated from the number of UMIs not detected for a given gene. The full length fragmented reads without xxx’ enrichment or UMI were aligned using RSEM against a reference transcriptome  which is depleted of mtRNA  rRNA  and pseudogenes as described above. The command ‘rsem calculate expression’ was used with parameters  p 4   paired end   seed length 24 for performing the alignment. The resulting gene levels in FPKM was used to build the expression matrix. Genome build: GRCz10 dr81 Supplementary files format and content: Tab delimited text files include gene levels UMI counts or FKPM for each sample.,Embryonic cell,,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for 5’ enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,Fish of each genotype were incrossed and embryos were collected. Embryos were dechorionated and allowed to grow at 28°C in standard fish water containing methylene blue till 50% epiboly stage.,developmental stage:50% epiboly|library preparation batch:p4|genotype:MZoep tz57,GSM2838392,GSM2838392: oep p4 S1; Danio rerio; RNA Seq,GSM2838392,,1,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for five prime enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina HiSeq 2500,,SRP123526,,,oep_p4_S1.R1.fastq.gz oep_p4_S1.R2.fastq.gz,fastq fastq,101140480.0,1806080.0,GSM2838392 r1,0:24 1:32,A:26745478;C:23576874;G:23869258;T:26866139;N:82731,24,32,,,26745478,23576874,23869258,26866139,82731,SRX3357228,SRS2656310,SRA627905,GEO,"Schier Lab, Molecular and Cellular Biology, Harvard University",2,0.80153,0.85401,0.14352,0.15536,0.85957,0.86048,0.7519,0.75548,24,32,B,B,biological fallback assumption,illumina,hiseq_era,full_length,poly_a,unknown,sc,single_cell_plate,smartseq,,United States,2017-11-02,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
44145,SRR6250305,SRX3357227,SRS2656309,SRP123526,PRJNA416939,Single cell RNAseq SMART seq2 of wild type TLAB and MZoep tz57 zebrafish embryos at 50% epiboly stage,GSE106466,Transcriptome Analysis,SMART seq2 was performed on single cells isolated from visually staged zebrafish embryos. Overall design: Samples were all sequenced in one batch. Some were generated with a five prime' UMI tagged method  and others are full length SMART seq2.,parent bioproject:PRJNA417291,pubmed:29700225,,oep p4 S19,GSM2838391,,tissue:Embryonic cell|developmental stage:50% epiboly|library preparation batch:p4|genotype:MZoep tz57,oep p4 S19,Bases 1 8 the UMI and template switch were trimmed from the remaining reads  trimmed stretches of polyA or polyT longer than 6bp and removed any reads that were shorter than 15 bp post trimming. We identified and depleted mtRNA and rRNA reads from the remaining reads by using Bowtie2 to align all reads against a transcriptome of mtRNA and rRNA  using the parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed   no unal. Remaining reads were then aligned to a modified version of the Ensembl Zv10 release 81 zebrafish transcriptome using Bowtie2 with parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed. The transcriptome start site of each gene was padded by 100bp to increase alignment rates  since reads were generated from the 5’ end of the transcript  and not all transcripts had correct annotations for their 5’ end. Additionally  pseudogenes were removed from the transcriptome as they sometimes interfered with real genes  given our 5’ coordinate selection described below. We chose the best alignment per read as the read with the most 5’ alignment coordinate to a transcript among those that aligned to the correct strand and had an alignment score of no more than 4 below the best alignment score overall. We then iteratively collapsed pairs of UMIs  if the sum of the minimum Phred quality scores across non matching bases was at most 30 to account for mis matched bases that were likely due to sequencing errors. A collapsed pair was represented for future comparisons by the UMI with a larger prior number of aligned reads unless the other UMI had at least 75% as many reads aligned and had equal or better UMI quality scores for those alignments evaluated as the softmin over aligned reads of the softmax over quality scores for the UMI bases in an aligned read. post collapsing  the number of transcripts observed per gene was calculated. This value was then adjusted based on the probability of a collision the drawing at random of the same UMI multiple times  which was calculated from the number of UMIs not detected for a given gene. The full length fragmented reads without xxx’ enrichment or UMI were aligned using RSEM against a reference transcriptome  which is depleted of mtRNA  rRNA  and pseudogenes as described above. The command ‘rsem calculate expression’ was used with parameters  p 4   paired end   seed length 24 for performing the alignment. The resulting gene levels in FPKM was used to build the expression matrix. Genome build: GRCz10 dr81 Supplementary files format and content: Tab delimited text files include gene levels UMI counts or FKPM for each sample.,Embryonic cell,,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for 5’ enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,Fish of each genotype were incrossed and embryos were collected. Embryos were dechorionated and allowed to grow at 28°C in standard fish water containing methylene blue till 50% epiboly stage.,developmental stage:50% epiboly|library preparation batch:p4|genotype:MZoep tz57,GSM2838391,GSM2838391: oep p4 S19; Danio rerio; RNA Seq,GSM2838391,,1,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for five prime enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina HiSeq 2500,,SRP123526,,,oep_p4_S19.R1.fastq.gz oep_p4_S19.R2.fastq.gz,fastq fastq,69556088.0,1242073.0,GSM2838391 r1,0:24 1:32,A:19403027;C:15217162;G:15475096;T:19405287;N:55516,24,32,,,19403027,15217162,15475096,19405287,55516,SRX3357227,SRS2656309,SRA627905,GEO,"Schier Lab, Molecular and Cellular Biology, Harvard University",2,0.76838,0.81461,0.07623,0.07663,0.85788,0.85811,0.73855,0.7503,24,32,B,B,biological fallback assumption,illumina,hiseq_era,full_length,poly_a,unknown,sc,single_cell_plate,smartseq,,United States,2017-11-02,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
44146,SRR6250304,SRX3357226,SRS2656308,SRP123526,PRJNA416939,Single cell RNAseq SMART seq2 of wild type TLAB and MZoep tz57 zebrafish embryos at 50% epiboly stage,GSE106466,Transcriptome Analysis,SMART seq2 was performed on single cells isolated from visually staged zebrafish embryos. Overall design: Samples were all sequenced in one batch. Some were generated with a five prime' UMI tagged method  and others are full length SMART seq2.,parent bioproject:PRJNA417291,pubmed:29700225,,oep p4 S18,GSM2838390,,tissue:Embryonic cell|developmental stage:50% epiboly|library preparation batch:p4|genotype:MZoep tz57,oep p4 S18,Bases 1 8 the UMI and template switch were trimmed from the remaining reads  trimmed stretches of polyA or polyT longer than 6bp and removed any reads that were shorter than 15 bp post trimming. We identified and depleted mtRNA and rRNA reads from the remaining reads by using Bowtie2 to align all reads against a transcriptome of mtRNA and rRNA  using the parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed   no unal. Remaining reads were then aligned to a modified version of the Ensembl Zv10 release 81 zebrafish transcriptome using Bowtie2 with parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed. The transcriptome start site of each gene was padded by 100bp to increase alignment rates  since reads were generated from the 5’ end of the transcript  and not all transcripts had correct annotations for their 5’ end. Additionally  pseudogenes were removed from the transcriptome as they sometimes interfered with real genes  given our 5’ coordinate selection described below. We chose the best alignment per read as the read with the most 5’ alignment coordinate to a transcript among those that aligned to the correct strand and had an alignment score of no more than 4 below the best alignment score overall. We then iteratively collapsed pairs of UMIs  if the sum of the minimum Phred quality scores across non matching bases was at most 30 to account for mis matched bases that were likely due to sequencing errors. A collapsed pair was represented for future comparisons by the UMI with a larger prior number of aligned reads unless the other UMI had at least 75% as many reads aligned and had equal or better UMI quality scores for those alignments evaluated as the softmin over aligned reads of the softmax over quality scores for the UMI bases in an aligned read. post collapsing  the number of transcripts observed per gene was calculated. This value was then adjusted based on the probability of a collision the drawing at random of the same UMI multiple times  which was calculated from the number of UMIs not detected for a given gene. The full length fragmented reads without xxx’ enrichment or UMI were aligned using RSEM against a reference transcriptome  which is depleted of mtRNA  rRNA  and pseudogenes as described above. The command ‘rsem calculate expression’ was used with parameters  p 4   paired end   seed length 24 for performing the alignment. The resulting gene levels in FPKM was used to build the expression matrix. Genome build: GRCz10 dr81 Supplementary files format and content: Tab delimited text files include gene levels UMI counts or FKPM for each sample.,Embryonic cell,,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for 5’ enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,Fish of each genotype were incrossed and embryos were collected. Embryos were dechorionated and allowed to grow at 28°C in standard fish water containing methylene blue till 50% epiboly stage.,developmental stage:50% epiboly|library preparation batch:p4|genotype:MZoep tz57,GSM2838390,GSM2838390: oep p4 S18; Danio rerio; RNA Seq,GSM2838390,,1,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for five prime enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina HiSeq 2500,,SRP123526,,,oep_p4_S18.R1.fastq.gz oep_p4_S18.R2.fastq.gz,fastq fastq,4183704.0,74709.0,GSM2838390 r1,,,,,,,,,,,,SRX3357226,SRS2656308,SRA627905,GEO,"Schier Lab, Molecular and Cellular Biology, Harvard University",2,0.77453,0.81404,0.07859,0.08027,0.90053,0.90011,0.77064,0.30009,24,32,B,B,biological fallback assumption,illumina,hiseq_era,full_length,poly_a,unknown,sc,single_cell_plate,smartseq,,United States,2017-11-02,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
44147,SRR6250303,SRX3357225,SRS2656307,SRP123526,PRJNA416939,Single cell RNAseq SMART seq2 of wild type TLAB and MZoep tz57 zebrafish embryos at 50% epiboly stage,GSE106466,Transcriptome Analysis,SMART seq2 was performed on single cells isolated from visually staged zebrafish embryos. Overall design: Samples were all sequenced in one batch. Some were generated with a five prime' UMI tagged method  and others are full length SMART seq2.,parent bioproject:PRJNA417291,pubmed:29700225,,oep p4 S17,GSM2838389,,tissue:Embryonic cell|developmental stage:50% epiboly|library preparation batch:p4|genotype:MZoep tz57,oep p4 S17,Bases 1 8 the UMI and template switch were trimmed from the remaining reads  trimmed stretches of polyA or polyT longer than 6bp and removed any reads that were shorter than 15 bp post trimming. We identified and depleted mtRNA and rRNA reads from the remaining reads by using Bowtie2 to align all reads against a transcriptome of mtRNA and rRNA  using the parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed   no unal. Remaining reads were then aligned to a modified version of the Ensembl Zv10 release 81 zebrafish transcriptome using Bowtie2 with parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed. The transcriptome start site of each gene was padded by 100bp to increase alignment rates  since reads were generated from the 5’ end of the transcript  and not all transcripts had correct annotations for their 5’ end. Additionally  pseudogenes were removed from the transcriptome as they sometimes interfered with real genes  given our 5’ coordinate selection described below. We chose the best alignment per read as the read with the most 5’ alignment coordinate to a transcript among those that aligned to the correct strand and had an alignment score of no more than 4 below the best alignment score overall. We then iteratively collapsed pairs of UMIs  if the sum of the minimum Phred quality scores across non matching bases was at most 30 to account for mis matched bases that were likely due to sequencing errors. A collapsed pair was represented for future comparisons by the UMI with a larger prior number of aligned reads unless the other UMI had at least 75% as many reads aligned and had equal or better UMI quality scores for those alignments evaluated as the softmin over aligned reads of the softmax over quality scores for the UMI bases in an aligned read. post collapsing  the number of transcripts observed per gene was calculated. This value was then adjusted based on the probability of a collision the drawing at random of the same UMI multiple times  which was calculated from the number of UMIs not detected for a given gene. The full length fragmented reads without xxx’ enrichment or UMI were aligned using RSEM against a reference transcriptome  which is depleted of mtRNA  rRNA  and pseudogenes as described above. The command ‘rsem calculate expression’ was used with parameters  p 4   paired end   seed length 24 for performing the alignment. The resulting gene levels in FPKM was used to build the expression matrix. Genome build: GRCz10 dr81 Supplementary files format and content: Tab delimited text files include gene levels UMI counts or FKPM for each sample.,Embryonic cell,,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for 5’ enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,Fish of each genotype were incrossed and embryos were collected. Embryos were dechorionated and allowed to grow at 28°C in standard fish water containing methylene blue till 50% epiboly stage.,developmental stage:50% epiboly|library preparation batch:p4|genotype:MZoep tz57,GSM2838389,GSM2838389: oep p4 S17; Danio rerio; RNA Seq,GSM2838389,,1,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for five prime enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina HiSeq 2500,,SRP123526,,,oep_p4_S17.R1.fastq.gz oep_p4_S17.R2.fastq.gz,fastq fastq,92090936.0,1644481.0,GSM2838389 r1,0:24 1:32,A:25372860;C:20478356;G:20907685;T:25257246;N:74789,24,32,,,25372860,20478356,20907685,25257246,74789,SRX3357225,SRS2656307,SRA627905,GEO,"Schier Lab, Molecular and Cellular Biology, Harvard University",2,0.78048,0.83088,0.06347,0.06502,0.84926,0.84916,0.73295,0.73519,24,32,B,B,biological fallback assumption,illumina,hiseq_era,full_length,poly_a,unknown,sc,single_cell_plate,smartseq,,United States,2017-11-02,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
44148,SRR6250302,SRX3357224,SRS2656306,SRP123526,PRJNA416939,Single cell RNAseq SMART seq2 of wild type TLAB and MZoep tz57 zebrafish embryos at 50% epiboly stage,GSE106466,Transcriptome Analysis,SMART seq2 was performed on single cells isolated from visually staged zebrafish embryos. Overall design: Samples were all sequenced in one batch. Some were generated with a five prime' UMI tagged method  and others are full length SMART seq2.,parent bioproject:PRJNA417291,pubmed:29700225,,oep p4 S16,GSM2838388,,tissue:Embryonic cell|developmental stage:50% epiboly|library preparation batch:p4|genotype:MZoep tz57,oep p4 S16,Bases 1 8 the UMI and template switch were trimmed from the remaining reads  trimmed stretches of polyA or polyT longer than 6bp and removed any reads that were shorter than 15 bp post trimming. We identified and depleted mtRNA and rRNA reads from the remaining reads by using Bowtie2 to align all reads against a transcriptome of mtRNA and rRNA  using the parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed   no unal. Remaining reads were then aligned to a modified version of the Ensembl Zv10 release 81 zebrafish transcriptome using Bowtie2 with parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed. The transcriptome start site of each gene was padded by 100bp to increase alignment rates  since reads were generated from the 5’ end of the transcript  and not all transcripts had correct annotations for their 5’ end. Additionally  pseudogenes were removed from the transcriptome as they sometimes interfered with real genes  given our 5’ coordinate selection described below. We chose the best alignment per read as the read with the most 5’ alignment coordinate to a transcript among those that aligned to the correct strand and had an alignment score of no more than 4 below the best alignment score overall. We then iteratively collapsed pairs of UMIs  if the sum of the minimum Phred quality scores across non matching bases was at most 30 to account for mis matched bases that were likely due to sequencing errors. A collapsed pair was represented for future comparisons by the UMI with a larger prior number of aligned reads unless the other UMI had at least 75% as many reads aligned and had equal or better UMI quality scores for those alignments evaluated as the softmin over aligned reads of the softmax over quality scores for the UMI bases in an aligned read. post collapsing  the number of transcripts observed per gene was calculated. This value was then adjusted based on the probability of a collision the drawing at random of the same UMI multiple times  which was calculated from the number of UMIs not detected for a given gene. The full length fragmented reads without xxx’ enrichment or UMI were aligned using RSEM against a reference transcriptome  which is depleted of mtRNA  rRNA  and pseudogenes as described above. The command ‘rsem calculate expression’ was used with parameters  p 4   paired end   seed length 24 for performing the alignment. The resulting gene levels in FPKM was used to build the expression matrix. Genome build: GRCz10 dr81 Supplementary files format and content: Tab delimited text files include gene levels UMI counts or FKPM for each sample.,Embryonic cell,,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for 5’ enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,Fish of each genotype were incrossed and embryos were collected. Embryos were dechorionated and allowed to grow at 28°C in standard fish water containing methylene blue till 50% epiboly stage.,developmental stage:50% epiboly|library preparation batch:p4|genotype:MZoep tz57,GSM2838388,GSM2838388: oep p4 S16; Danio rerio; RNA Seq,GSM2838388,,1,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for five prime enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina HiSeq 2500,,SRP123526,,,oep_p4_S16.R1.fastq.gz oep_p4_S16.R2.fastq.gz,fastq fastq,78362088.0,1399323.0,GSM2838388 r1,0:24 1:32,A:21627730;C:17366835;G:17737716;T:21567579;N:62228,24,32,,,21627730,17366835,17737716,21567579,62228,SRX3357224,SRS2656306,SRA627905,GEO,"Schier Lab, Molecular and Cellular Biology, Harvard University",2,0.76846,0.81735,0.0843,0.08632,0.85851,0.85847,0.74942,0.75016,24,32,B,B,biological fallback assumption,illumina,hiseq_era,full_length,poly_a,unknown,sc,single_cell_plate,smartseq,,United States,2017-11-02,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
44149,SRR6250301,SRX3357223,SRS2656305,SRP123526,PRJNA416939,Single cell RNAseq SMART seq2 of wild type TLAB and MZoep tz57 zebrafish embryos at 50% epiboly stage,GSE106466,Transcriptome Analysis,SMART seq2 was performed on single cells isolated from visually staged zebrafish embryos. Overall design: Samples were all sequenced in one batch. Some were generated with a five prime' UMI tagged method  and others are full length SMART seq2.,parent bioproject:PRJNA417291,pubmed:29700225,,oep p4 S15,GSM2838387,,tissue:Embryonic cell|developmental stage:50% epiboly|library preparation batch:p4|genotype:MZoep tz57,oep p4 S15,Bases 1 8 the UMI and template switch were trimmed from the remaining reads  trimmed stretches of polyA or polyT longer than 6bp and removed any reads that were shorter than 15 bp post trimming. We identified and depleted mtRNA and rRNA reads from the remaining reads by using Bowtie2 to align all reads against a transcriptome of mtRNA and rRNA  using the parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed   no unal. Remaining reads were then aligned to a modified version of the Ensembl Zv10 release 81 zebrafish transcriptome using Bowtie2 with parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed. The transcriptome start site of each gene was padded by 100bp to increase alignment rates  since reads were generated from the 5’ end of the transcript  and not all transcripts had correct annotations for their 5’ end. Additionally  pseudogenes were removed from the transcriptome as they sometimes interfered with real genes  given our 5’ coordinate selection described below. We chose the best alignment per read as the read with the most 5’ alignment coordinate to a transcript among those that aligned to the correct strand and had an alignment score of no more than 4 below the best alignment score overall. We then iteratively collapsed pairs of UMIs  if the sum of the minimum Phred quality scores across non matching bases was at most 30 to account for mis matched bases that were likely due to sequencing errors. A collapsed pair was represented for future comparisons by the UMI with a larger prior number of aligned reads unless the other UMI had at least 75% as many reads aligned and had equal or better UMI quality scores for those alignments evaluated as the softmin over aligned reads of the softmax over quality scores for the UMI bases in an aligned read. post collapsing  the number of transcripts observed per gene was calculated. This value was then adjusted based on the probability of a collision the drawing at random of the same UMI multiple times  which was calculated from the number of UMIs not detected for a given gene. The full length fragmented reads without xxx’ enrichment or UMI were aligned using RSEM against a reference transcriptome  which is depleted of mtRNA  rRNA  and pseudogenes as described above. The command ‘rsem calculate expression’ was used with parameters  p 4   paired end   seed length 24 for performing the alignment. The resulting gene levels in FPKM was used to build the expression matrix. Genome build: GRCz10 dr81 Supplementary files format and content: Tab delimited text files include gene levels UMI counts or FKPM for each sample.,Embryonic cell,,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for 5’ enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,Fish of each genotype were incrossed and embryos were collected. Embryos were dechorionated and allowed to grow at 28°C in standard fish water containing methylene blue till 50% epiboly stage.,developmental stage:50% epiboly|library preparation batch:p4|genotype:MZoep tz57,GSM2838387,GSM2838387: oep p4 S15; Danio rerio; RNA Seq,GSM2838387,,1,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for five prime enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina HiSeq 2500,,SRP123526,,,oep_p4_S15.R1.fastq.gz oep_p4_S15.R2.fastq.gz,fastq fastq,3443832.0,61497.0,GSM2838387 r1,0:24 1:32,A:899084;C:804618;G:810086;T:927528;N:2516,24,32,,,899084,804618,810086,927528,2516,SRX3357223,SRS2656305,SRA627905,GEO,"Schier Lab, Molecular and Cellular Biology, Harvard University",2,0.77278,0.83379,0.18796,0.2009,0.90871,0.90514,0.34291,0.71589,24,32,B,B,biological fallback assumption,illumina,hiseq_era,full_length,poly_a,unknown,sc,single_cell_plate,smartseq,,United States,2017-11-02,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
44150,SRR6250300,SRX3357222,SRS2656304,SRP123526,PRJNA416939,Single cell RNAseq SMART seq2 of wild type TLAB and MZoep tz57 zebrafish embryos at 50% epiboly stage,GSE106466,Transcriptome Analysis,SMART seq2 was performed on single cells isolated from visually staged zebrafish embryos. Overall design: Samples were all sequenced in one batch. Some were generated with a five prime' UMI tagged method  and others are full length SMART seq2.,parent bioproject:PRJNA417291,pubmed:29700225,,oep p4 S14,GSM2838386,,tissue:Embryonic cell|developmental stage:50% epiboly|library preparation batch:p4|genotype:MZoep tz57,oep p4 S14,Bases 1 8 the UMI and template switch were trimmed from the remaining reads  trimmed stretches of polyA or polyT longer than 6bp and removed any reads that were shorter than 15 bp post trimming. We identified and depleted mtRNA and rRNA reads from the remaining reads by using Bowtie2 to align all reads against a transcriptome of mtRNA and rRNA  using the parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed   no unal. Remaining reads were then aligned to a modified version of the Ensembl Zv10 release 81 zebrafish transcriptome using Bowtie2 with parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed. The transcriptome start site of each gene was padded by 100bp to increase alignment rates  since reads were generated from the 5’ end of the transcript  and not all transcripts had correct annotations for their 5’ end. Additionally  pseudogenes were removed from the transcriptome as they sometimes interfered with real genes  given our 5’ coordinate selection described below. We chose the best alignment per read as the read with the most 5’ alignment coordinate to a transcript among those that aligned to the correct strand and had an alignment score of no more than 4 below the best alignment score overall. We then iteratively collapsed pairs of UMIs  if the sum of the minimum Phred quality scores across non matching bases was at most 30 to account for mis matched bases that were likely due to sequencing errors. A collapsed pair was represented for future comparisons by the UMI with a larger prior number of aligned reads unless the other UMI had at least 75% as many reads aligned and had equal or better UMI quality scores for those alignments evaluated as the softmin over aligned reads of the softmax over quality scores for the UMI bases in an aligned read. post collapsing  the number of transcripts observed per gene was calculated. This value was then adjusted based on the probability of a collision the drawing at random of the same UMI multiple times  which was calculated from the number of UMIs not detected for a given gene. The full length fragmented reads without xxx’ enrichment or UMI were aligned using RSEM against a reference transcriptome  which is depleted of mtRNA  rRNA  and pseudogenes as described above. The command ‘rsem calculate expression’ was used with parameters  p 4   paired end   seed length 24 for performing the alignment. The resulting gene levels in FPKM was used to build the expression matrix. Genome build: GRCz10 dr81 Supplementary files format and content: Tab delimited text files include gene levels UMI counts or FKPM for each sample.,Embryonic cell,,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for 5’ enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,Fish of each genotype were incrossed and embryos were collected. Embryos were dechorionated and allowed to grow at 28°C in standard fish water containing methylene blue till 50% epiboly stage.,developmental stage:50% epiboly|library preparation batch:p4|genotype:MZoep tz57,GSM2838386,GSM2838386: oep p4 S14; Danio rerio; RNA Seq,GSM2838386,,1,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for five prime enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina HiSeq 2500,,SRP123526,,,oep_p4_S14.R1.fastq.gz oep_p4_S14.R2.fastq.gz,fastq fastq,84153440.0,1502740.0,GSM2838386 r1,0:24 1:32,A:22314812;C:19454797;G:19684266;T:22631594;N:67971,24,32,,,22314812,19454797,19684266,22631594,67971,SRX3357222,SRS2656304,SRA627905,GEO,"Schier Lab, Molecular and Cellular Biology, Harvard University",2,0.78868,0.84021,0.15916,0.17058,0.84327,0.8434,0.71882,0.72779,24,32,B,B,biological fallback assumption,illumina,hiseq_era,full_length,poly_a,unknown,sc,single_cell_plate,smartseq,,United States,2017-11-02,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
44151,SRR6250299,SRX3357221,SRS2656303,SRP123526,PRJNA416939,Single cell RNAseq SMART seq2 of wild type TLAB and MZoep tz57 zebrafish embryos at 50% epiboly stage,GSE106466,Transcriptome Analysis,SMART seq2 was performed on single cells isolated from visually staged zebrafish embryos. Overall design: Samples were all sequenced in one batch. Some were generated with a five prime' UMI tagged method  and others are full length SMART seq2.,parent bioproject:PRJNA417291,pubmed:29700225,,oep p4 S13,GSM2838385,,tissue:Embryonic cell|developmental stage:50% epiboly|library preparation batch:p4|genotype:MZoep tz57,oep p4 S13,Bases 1 8 the UMI and template switch were trimmed from the remaining reads  trimmed stretches of polyA or polyT longer than 6bp and removed any reads that were shorter than 15 bp post trimming. We identified and depleted mtRNA and rRNA reads from the remaining reads by using Bowtie2 to align all reads against a transcriptome of mtRNA and rRNA  using the parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed   no unal. Remaining reads were then aligned to a modified version of the Ensembl Zv10 release 81 zebrafish transcriptome using Bowtie2 with parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed. The transcriptome start site of each gene was padded by 100bp to increase alignment rates  since reads were generated from the 5’ end of the transcript  and not all transcripts had correct annotations for their 5’ end. Additionally  pseudogenes were removed from the transcriptome as they sometimes interfered with real genes  given our 5’ coordinate selection described below. We chose the best alignment per read as the read with the most 5’ alignment coordinate to a transcript among those that aligned to the correct strand and had an alignment score of no more than 4 below the best alignment score overall. We then iteratively collapsed pairs of UMIs  if the sum of the minimum Phred quality scores across non matching bases was at most 30 to account for mis matched bases that were likely due to sequencing errors. A collapsed pair was represented for future comparisons by the UMI with a larger prior number of aligned reads unless the other UMI had at least 75% as many reads aligned and had equal or better UMI quality scores for those alignments evaluated as the softmin over aligned reads of the softmax over quality scores for the UMI bases in an aligned read. post collapsing  the number of transcripts observed per gene was calculated. This value was then adjusted based on the probability of a collision the drawing at random of the same UMI multiple times  which was calculated from the number of UMIs not detected for a given gene. The full length fragmented reads without xxx’ enrichment or UMI were aligned using RSEM against a reference transcriptome  which is depleted of mtRNA  rRNA  and pseudogenes as described above. The command ‘rsem calculate expression’ was used with parameters  p 4   paired end   seed length 24 for performing the alignment. The resulting gene levels in FPKM was used to build the expression matrix. Genome build: GRCz10 dr81 Supplementary files format and content: Tab delimited text files include gene levels UMI counts or FKPM for each sample.,Embryonic cell,,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for 5’ enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,Fish of each genotype were incrossed and embryos were collected. Embryos were dechorionated and allowed to grow at 28°C in standard fish water containing methylene blue till 50% epiboly stage.,developmental stage:50% epiboly|library preparation batch:p4|genotype:MZoep tz57,GSM2838385,GSM2838385: oep p4 S13; Danio rerio; RNA Seq,GSM2838385,,1,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for five prime enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina HiSeq 2500,,SRP123526,,,oep_p4_S13.R1.fastq.gz oep_p4_S13.R2.fastq.gz,fastq fastq,65871120.0,1176270.0,GSM2838385 r1,0:24 1:32,A:17192209;C:15459137;G:15653476;T:17512736;N:53562,24,32,,,17192209,15459137,15653476,17512736,53562,SRX3357221,SRS2656303,SRA627905,GEO,"Schier Lab, Molecular and Cellular Biology, Harvard University",2,0.78914,0.84191,0.19002,0.20308,0.84088,0.83968,0.67593,0.67633,24,32,B,B,biological fallback assumption,illumina,hiseq_era,full_length,poly_a,unknown,sc,single_cell_plate,smartseq,,United States,2017-11-02,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
44152,SRR6250298,SRX3357220,SRS2656302,SRP123526,PRJNA416939,Single cell RNAseq SMART seq2 of wild type TLAB and MZoep tz57 zebrafish embryos at 50% epiboly stage,GSE106466,Transcriptome Analysis,SMART seq2 was performed on single cells isolated from visually staged zebrafish embryos. Overall design: Samples were all sequenced in one batch. Some were generated with a five prime' UMI tagged method  and others are full length SMART seq2.,parent bioproject:PRJNA417291,pubmed:29700225,,oep p4 S12,GSM2838384,,tissue:Embryonic cell|developmental stage:50% epiboly|library preparation batch:p4|genotype:MZoep tz57,oep p4 S12,Bases 1 8 the UMI and template switch were trimmed from the remaining reads  trimmed stretches of polyA or polyT longer than 6bp and removed any reads that were shorter than 15 bp post trimming. We identified and depleted mtRNA and rRNA reads from the remaining reads by using Bowtie2 to align all reads against a transcriptome of mtRNA and rRNA  using the parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed   no unal. Remaining reads were then aligned to a modified version of the Ensembl Zv10 release 81 zebrafish transcriptome using Bowtie2 with parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed. The transcriptome start site of each gene was padded by 100bp to increase alignment rates  since reads were generated from the 5’ end of the transcript  and not all transcripts had correct annotations for their 5’ end. Additionally  pseudogenes were removed from the transcriptome as they sometimes interfered with real genes  given our 5’ coordinate selection described below. We chose the best alignment per read as the read with the most 5’ alignment coordinate to a transcript among those that aligned to the correct strand and had an alignment score of no more than 4 below the best alignment score overall. We then iteratively collapsed pairs of UMIs  if the sum of the minimum Phred quality scores across non matching bases was at most 30 to account for mis matched bases that were likely due to sequencing errors. A collapsed pair was represented for future comparisons by the UMI with a larger prior number of aligned reads unless the other UMI had at least 75% as many reads aligned and had equal or better UMI quality scores for those alignments evaluated as the softmin over aligned reads of the softmax over quality scores for the UMI bases in an aligned read. post collapsing  the number of transcripts observed per gene was calculated. This value was then adjusted based on the probability of a collision the drawing at random of the same UMI multiple times  which was calculated from the number of UMIs not detected for a given gene. The full length fragmented reads without xxx’ enrichment or UMI were aligned using RSEM against a reference transcriptome  which is depleted of mtRNA  rRNA  and pseudogenes as described above. The command ‘rsem calculate expression’ was used with parameters  p 4   paired end   seed length 24 for performing the alignment. The resulting gene levels in FPKM was used to build the expression matrix. Genome build: GRCz10 dr81 Supplementary files format and content: Tab delimited text files include gene levels UMI counts or FKPM for each sample.,Embryonic cell,,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for 5’ enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,Fish of each genotype were incrossed and embryos were collected. Embryos were dechorionated and allowed to grow at 28°C in standard fish water containing methylene blue till 50% epiboly stage.,developmental stage:50% epiboly|library preparation batch:p4|genotype:MZoep tz57,GSM2838384,GSM2838384: oep p4 S12; Danio rerio; RNA Seq,GSM2838384,,1,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for five prime enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina HiSeq 2500,,SRP123526,,,oep_p4_S12.R1.fastq.gz oep_p4_S12.R2.fastq.gz,fastq fastq,75987520.0,1356920.0,GSM2838384 r1,0:24 1:32,A:20433443;C:17367926;G:17661569;T:20460743;N:63839,24,32,,,20433443,17367926,17661569,20460743,63839,SRX3357220,SRS2656302,SRA627905,GEO,"Schier Lab, Molecular and Cellular Biology, Harvard University",2,0.78514,0.83059,0.14453,0.15378,0.87091,0.87081,0.72523,0.72866,24,32,B,B,biological fallback assumption,illumina,hiseq_era,full_length,poly_a,unknown,sc,single_cell_plate,smartseq,,United States,2017-11-02,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
44153,SRR6250297,SRX3357219,SRS2656301,SRP123526,PRJNA416939,Single cell RNAseq SMART seq2 of wild type TLAB and MZoep tz57 zebrafish embryos at 50% epiboly stage,GSE106466,Transcriptome Analysis,SMART seq2 was performed on single cells isolated from visually staged zebrafish embryos. Overall design: Samples were all sequenced in one batch. Some were generated with a five prime' UMI tagged method  and others are full length SMART seq2.,parent bioproject:PRJNA417291,pubmed:29700225,,oep p4 S11,GSM2838383,,tissue:Embryonic cell|developmental stage:50% epiboly|library preparation batch:p4|genotype:MZoep tz57,oep p4 S11,Bases 1 8 the UMI and template switch were trimmed from the remaining reads  trimmed stretches of polyA or polyT longer than 6bp and removed any reads that were shorter than 15 bp post trimming. We identified and depleted mtRNA and rRNA reads from the remaining reads by using Bowtie2 to align all reads against a transcriptome of mtRNA and rRNA  using the parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed   no unal. Remaining reads were then aligned to a modified version of the Ensembl Zv10 release 81 zebrafish transcriptome using Bowtie2 with parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed. The transcriptome start site of each gene was padded by 100bp to increase alignment rates  since reads were generated from the 5’ end of the transcript  and not all transcripts had correct annotations for their 5’ end. Additionally  pseudogenes were removed from the transcriptome as they sometimes interfered with real genes  given our 5’ coordinate selection described below. We chose the best alignment per read as the read with the most 5’ alignment coordinate to a transcript among those that aligned to the correct strand and had an alignment score of no more than 4 below the best alignment score overall. We then iteratively collapsed pairs of UMIs  if the sum of the minimum Phred quality scores across non matching bases was at most 30 to account for mis matched bases that were likely due to sequencing errors. A collapsed pair was represented for future comparisons by the UMI with a larger prior number of aligned reads unless the other UMI had at least 75% as many reads aligned and had equal or better UMI quality scores for those alignments evaluated as the softmin over aligned reads of the softmax over quality scores for the UMI bases in an aligned read. post collapsing  the number of transcripts observed per gene was calculated. This value was then adjusted based on the probability of a collision the drawing at random of the same UMI multiple times  which was calculated from the number of UMIs not detected for a given gene. The full length fragmented reads without xxx’ enrichment or UMI were aligned using RSEM against a reference transcriptome  which is depleted of mtRNA  rRNA  and pseudogenes as described above. The command ‘rsem calculate expression’ was used with parameters  p 4   paired end   seed length 24 for performing the alignment. The resulting gene levels in FPKM was used to build the expression matrix. Genome build: GRCz10 dr81 Supplementary files format and content: Tab delimited text files include gene levels UMI counts or FKPM for each sample.,Embryonic cell,,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for 5’ enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,Fish of each genotype were incrossed and embryos were collected. Embryos were dechorionated and allowed to grow at 28°C in standard fish water containing methylene blue till 50% epiboly stage.,developmental stage:50% epiboly|library preparation batch:p4|genotype:MZoep tz57,GSM2838383,GSM2838383: oep p4 S11; Danio rerio; RNA Seq,GSM2838383,,1,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for five prime enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina HiSeq 2500,,SRP123526,,,oep_p4_S11.R1.fastq.gz oep_p4_S11.R2.fastq.gz,fastq fastq,3793384.0,67739.0,GSM2838383 r1,0:24 1:32,A:1003375;C:874215;G:890420;T:1022315;N:3059,24,32,,,1003375,874215,890420,1022315,3059,SRX3357219,SRS2656301,SRA627905,GEO,"Schier Lab, Molecular and Cellular Biology, Harvard University",2,0.50843,0.53703,0.11615,0.12129,0.92685,0.926,0.75144,0.74516,24,32,B,B,biological fallback assumption,illumina,hiseq_era,full_length,poly_a,unknown,sc,single_cell_plate,smartseq,,United States,2017-11-02,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
44154,SRR6250296,SRX3357218,SRS2656300,SRP123526,PRJNA416939,Single cell RNAseq SMART seq2 of wild type TLAB and MZoep tz57 zebrafish embryos at 50% epiboly stage,GSE106466,Transcriptome Analysis,SMART seq2 was performed on single cells isolated from visually staged zebrafish embryos. Overall design: Samples were all sequenced in one batch. Some were generated with a five prime' UMI tagged method  and others are full length SMART seq2.,parent bioproject:PRJNA417291,pubmed:29700225,,oep p4 S10,GSM2838382,,tissue:Embryonic cell|developmental stage:50% epiboly|library preparation batch:p4|genotype:MZoep tz57,oep p4 S10,Bases 1 8 the UMI and template switch were trimmed from the remaining reads  trimmed stretches of polyA or polyT longer than 6bp and removed any reads that were shorter than 15 bp post trimming. We identified and depleted mtRNA and rRNA reads from the remaining reads by using Bowtie2 to align all reads against a transcriptome of mtRNA and rRNA  using the parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed   no unal. Remaining reads were then aligned to a modified version of the Ensembl Zv10 release 81 zebrafish transcriptome using Bowtie2 with parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed. The transcriptome start site of each gene was padded by 100bp to increase alignment rates  since reads were generated from the 5’ end of the transcript  and not all transcripts had correct annotations for their 5’ end. Additionally  pseudogenes were removed from the transcriptome as they sometimes interfered with real genes  given our 5’ coordinate selection described below. We chose the best alignment per read as the read with the most 5’ alignment coordinate to a transcript among those that aligned to the correct strand and had an alignment score of no more than 4 below the best alignment score overall. We then iteratively collapsed pairs of UMIs  if the sum of the minimum Phred quality scores across non matching bases was at most 30 to account for mis matched bases that were likely due to sequencing errors. A collapsed pair was represented for future comparisons by the UMI with a larger prior number of aligned reads unless the other UMI had at least 75% as many reads aligned and had equal or better UMI quality scores for those alignments evaluated as the softmin over aligned reads of the softmax over quality scores for the UMI bases in an aligned read. post collapsing  the number of transcripts observed per gene was calculated. This value was then adjusted based on the probability of a collision the drawing at random of the same UMI multiple times  which was calculated from the number of UMIs not detected for a given gene. The full length fragmented reads without xxx’ enrichment or UMI were aligned using RSEM against a reference transcriptome  which is depleted of mtRNA  rRNA  and pseudogenes as described above. The command ‘rsem calculate expression’ was used with parameters  p 4   paired end   seed length 24 for performing the alignment. The resulting gene levels in FPKM was used to build the expression matrix. Genome build: GRCz10 dr81 Supplementary files format and content: Tab delimited text files include gene levels UMI counts or FKPM for each sample.,Embryonic cell,,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for 5’ enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,Fish of each genotype were incrossed and embryos were collected. Embryos were dechorionated and allowed to grow at 28°C in standard fish water containing methylene blue till 50% epiboly stage.,developmental stage:50% epiboly|library preparation batch:p4|genotype:MZoep tz57,GSM2838382,GSM2838382: oep p4 S10; Danio rerio; RNA Seq,GSM2838382,,1,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for five prime enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina HiSeq 2500,,SRP123526,,,oep_p4_S10.R1.fastq.gz oep_p4_S10.R2.fastq.gz,fastq fastq,67491256.0,1205201.0,GSM2838382 r1,0:24 1:32,A:18217694;C:15400421;G:15669743;T:18150401;N:52997,24,32,,,18217694,15400421,15669743,18150401,52997,SRX3357218,SRS2656300,SRA627905,GEO,"Schier Lab, Molecular and Cellular Biology, Harvard University",2,0.7981,0.84281,0.13672,0.14575,0.87411,0.87334,0.76971,0.77541,24,32,B,B,biological fallback assumption,illumina,hiseq_era,full_length,poly_a,unknown,sc,single_cell_plate,smartseq,,United States,2017-11-02,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
44155,SRR6250295,SRX3357217,SRS2656299,SRP123526,PRJNA416939,Single cell RNAseq SMART seq2 of wild type TLAB and MZoep tz57 zebrafish embryos at 50% epiboly stage,GSE106466,Transcriptome Analysis,SMART seq2 was performed on single cells isolated from visually staged zebrafish embryos. Overall design: Samples were all sequenced in one batch. Some were generated with a five prime' UMI tagged method  and others are full length SMART seq2.,parent bioproject:PRJNA417291,pubmed:29700225,,oep p3 S9,GSM2838381,,tissue:Embryonic cell|developmental stage:50% epiboly|library preparation batch:p3|genotype:MZoep tz57,oep p3 S9,Bases 1 8 the UMI and template switch were trimmed from the remaining reads  trimmed stretches of polyA or polyT longer than 6bp and removed any reads that were shorter than 15 bp post trimming. We identified and depleted mtRNA and rRNA reads from the remaining reads by using Bowtie2 to align all reads against a transcriptome of mtRNA and rRNA  using the parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed   no unal. Remaining reads were then aligned to a modified version of the Ensembl Zv10 release 81 zebrafish transcriptome using Bowtie2 with parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed. The transcriptome start site of each gene was padded by 100bp to increase alignment rates  since reads were generated from the 5’ end of the transcript  and not all transcripts had correct annotations for their 5’ end. Additionally  pseudogenes were removed from the transcriptome as they sometimes interfered with real genes  given our 5’ coordinate selection described below. We chose the best alignment per read as the read with the most 5’ alignment coordinate to a transcript among those that aligned to the correct strand and had an alignment score of no more than 4 below the best alignment score overall. We then iteratively collapsed pairs of UMIs  if the sum of the minimum Phred quality scores across non matching bases was at most 30 to account for mis matched bases that were likely due to sequencing errors. A collapsed pair was represented for future comparisons by the UMI with a larger prior number of aligned reads unless the other UMI had at least 75% as many reads aligned and had equal or better UMI quality scores for those alignments evaluated as the softmin over aligned reads of the softmax over quality scores for the UMI bases in an aligned read. post collapsing  the number of transcripts observed per gene was calculated. This value was then adjusted based on the probability of a collision the drawing at random of the same UMI multiple times  which was calculated from the number of UMIs not detected for a given gene. The full length fragmented reads without xxx’ enrichment or UMI were aligned using RSEM against a reference transcriptome  which is depleted of mtRNA  rRNA  and pseudogenes as described above. The command ‘rsem calculate expression’ was used with parameters  p 4   paired end   seed length 24 for performing the alignment. The resulting gene levels in FPKM was used to build the expression matrix. Genome build: GRCz10 dr81 Supplementary files format and content: Tab delimited text files include gene levels UMI counts or FKPM for each sample.,Embryonic cell,,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for 5’ enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,Fish of each genotype were incrossed and embryos were collected. Embryos were dechorionated and allowed to grow at 28°C in standard fish water containing methylene blue till 50% epiboly stage.,developmental stage:50% epiboly|library preparation batch:p3|genotype:MZoep tz57,GSM2838381,GSM2838381: oep p3 S9; Danio rerio; RNA Seq,GSM2838381,,1,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for five prime enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina HiSeq 2500,,SRP123526,,,oep_p3_S9.R1.fastq.gz oep_p3_S9.R2.fastq.gz,fastq fastq,66837680.0,1193530.0,GSM2838381 r1,0:24 1:32,A:17920696;C:15328550;G:15593377;T:17941985;N:53072,24,32,,,17920696,15328550,15593377,17941985,53072,SRX3357217,SRS2656299,SRA627905,GEO,"Schier Lab, Molecular and Cellular Biology, Harvard University",2,0.80301,0.85076,0.13595,0.14553,0.85876,0.8578,0.70612,0.70991,24,32,B,B,biological fallback assumption,illumina,hiseq_era,full_length,poly_a,unknown,sc,single_cell_plate,smartseq,,United States,2017-11-02,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
44156,SRR6250294,SRX3357216,SRS2656297,SRP123526,PRJNA416939,Single cell RNAseq SMART seq2 of wild type TLAB and MZoep tz57 zebrafish embryos at 50% epiboly stage,GSE106466,Transcriptome Analysis,SMART seq2 was performed on single cells isolated from visually staged zebrafish embryos. Overall design: Samples were all sequenced in one batch. Some were generated with a five prime' UMI tagged method  and others are full length SMART seq2.,parent bioproject:PRJNA417291,pubmed:29700225,,oep p3 S8,GSM2838380,,tissue:Embryonic cell|developmental stage:50% epiboly|library preparation batch:p3|genotype:MZoep tz57,oep p3 S8,Bases 1 8 the UMI and template switch were trimmed from the remaining reads  trimmed stretches of polyA or polyT longer than 6bp and removed any reads that were shorter than 15 bp post trimming. We identified and depleted mtRNA and rRNA reads from the remaining reads by using Bowtie2 to align all reads against a transcriptome of mtRNA and rRNA  using the parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed   no unal. Remaining reads were then aligned to a modified version of the Ensembl Zv10 release 81 zebrafish transcriptome using Bowtie2 with parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed. The transcriptome start site of each gene was padded by 100bp to increase alignment rates  since reads were generated from the 5’ end of the transcript  and not all transcripts had correct annotations for their 5’ end. Additionally  pseudogenes were removed from the transcriptome as they sometimes interfered with real genes  given our 5’ coordinate selection described below. We chose the best alignment per read as the read with the most 5’ alignment coordinate to a transcript among those that aligned to the correct strand and had an alignment score of no more than 4 below the best alignment score overall. We then iteratively collapsed pairs of UMIs  if the sum of the minimum Phred quality scores across non matching bases was at most 30 to account for mis matched bases that were likely due to sequencing errors. A collapsed pair was represented for future comparisons by the UMI with a larger prior number of aligned reads unless the other UMI had at least 75% as many reads aligned and had equal or better UMI quality scores for those alignments evaluated as the softmin over aligned reads of the softmax over quality scores for the UMI bases in an aligned read. post collapsing  the number of transcripts observed per gene was calculated. This value was then adjusted based on the probability of a collision the drawing at random of the same UMI multiple times  which was calculated from the number of UMIs not detected for a given gene. The full length fragmented reads without xxx’ enrichment or UMI were aligned using RSEM against a reference transcriptome  which is depleted of mtRNA  rRNA  and pseudogenes as described above. The command ‘rsem calculate expression’ was used with parameters  p 4   paired end   seed length 24 for performing the alignment. The resulting gene levels in FPKM was used to build the expression matrix. Genome build: GRCz10 dr81 Supplementary files format and content: Tab delimited text files include gene levels UMI counts or FKPM for each sample.,Embryonic cell,,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for 5’ enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,Fish of each genotype were incrossed and embryos were collected. Embryos were dechorionated and allowed to grow at 28°C in standard fish water containing methylene blue till 50% epiboly stage.,developmental stage:50% epiboly|library preparation batch:p3|genotype:MZoep tz57,GSM2838380,GSM2838380: oep p3 S8; Danio rerio; RNA Seq,GSM2838380,,1,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for five prime enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina HiSeq 2500,,SRP123526,,,oep_p3_S8.R1.fastq.gz oep_p3_S8.R2.fastq.gz,fastq fastq,83751080.0,1495555.0,GSM2838380 r1,0:24 1:32,A:22095926;C:19582224;G:19786171;T:22219235;N:67524,24,32,,,22095926,19582224,19786171,22219235,67524,SRX3357216,SRS2656297,SRA627905,GEO,"Schier Lab, Molecular and Cellular Biology, Harvard University",2,0.8005,0.85154,0.1633,0.17653,0.84461,0.84465,0.74525,0.74414,24,32,B,B,biological fallback assumption,illumina,hiseq_era,full_length,poly_a,unknown,sc,single_cell_plate,smartseq,,United States,2017-11-02,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
44157,SRR6250293,SRX3357215,SRS2656298,SRP123526,PRJNA416939,Single cell RNAseq SMART seq2 of wild type TLAB and MZoep tz57 zebrafish embryos at 50% epiboly stage,GSE106466,Transcriptome Analysis,SMART seq2 was performed on single cells isolated from visually staged zebrafish embryos. Overall design: Samples were all sequenced in one batch. Some were generated with a five prime' UMI tagged method  and others are full length SMART seq2.,parent bioproject:PRJNA417291,pubmed:29700225,,oep p3 S7,GSM2838379,,tissue:Embryonic cell|developmental stage:50% epiboly|library preparation batch:p3|genotype:MZoep tz57,oep p3 S7,Bases 1 8 the UMI and template switch were trimmed from the remaining reads  trimmed stretches of polyA or polyT longer than 6bp and removed any reads that were shorter than 15 bp post trimming. We identified and depleted mtRNA and rRNA reads from the remaining reads by using Bowtie2 to align all reads against a transcriptome of mtRNA and rRNA  using the parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed   no unal. Remaining reads were then aligned to a modified version of the Ensembl Zv10 release 81 zebrafish transcriptome using Bowtie2 with parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed. The transcriptome start site of each gene was padded by 100bp to increase alignment rates  since reads were generated from the 5’ end of the transcript  and not all transcripts had correct annotations for their 5’ end. Additionally  pseudogenes were removed from the transcriptome as they sometimes interfered with real genes  given our 5’ coordinate selection described below. We chose the best alignment per read as the read with the most 5’ alignment coordinate to a transcript among those that aligned to the correct strand and had an alignment score of no more than 4 below the best alignment score overall. We then iteratively collapsed pairs of UMIs  if the sum of the minimum Phred quality scores across non matching bases was at most 30 to account for mis matched bases that were likely due to sequencing errors. A collapsed pair was represented for future comparisons by the UMI with a larger prior number of aligned reads unless the other UMI had at least 75% as many reads aligned and had equal or better UMI quality scores for those alignments evaluated as the softmin over aligned reads of the softmax over quality scores for the UMI bases in an aligned read. post collapsing  the number of transcripts observed per gene was calculated. This value was then adjusted based on the probability of a collision the drawing at random of the same UMI multiple times  which was calculated from the number of UMIs not detected for a given gene. The full length fragmented reads without xxx’ enrichment or UMI were aligned using RSEM against a reference transcriptome  which is depleted of mtRNA  rRNA  and pseudogenes as described above. The command ‘rsem calculate expression’ was used with parameters  p 4   paired end   seed length 24 for performing the alignment. The resulting gene levels in FPKM was used to build the expression matrix. Genome build: GRCz10 dr81 Supplementary files format and content: Tab delimited text files include gene levels UMI counts or FKPM for each sample.,Embryonic cell,,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for 5’ enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,Fish of each genotype were incrossed and embryos were collected. Embryos were dechorionated and allowed to grow at 28°C in standard fish water containing methylene blue till 50% epiboly stage.,developmental stage:50% epiboly|library preparation batch:p3|genotype:MZoep tz57,GSM2838379,GSM2838379: oep p3 S7; Danio rerio; RNA Seq,GSM2838379,,1,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for five prime enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina HiSeq 2500,,SRP123526,,,oep_p3_S7.R1.fastq.gz oep_p3_S7.R2.fastq.gz,fastq fastq,74253536.0,1325956.0,GSM2838379 r1,0:24 1:32,A:19569753;C:17295764;G:17554585;T:19773058;N:60376,24,32,,,19569753,17295764,17554585,19773058,60376,SRX3357215,SRS2656298,SRA627905,GEO,"Schier Lab, Molecular and Cellular Biology, Harvard University",2,0.79969,0.85144,0.16576,0.17848,0.85003,0.84873,0.72639,0.72085,24,32,B,B,biological fallback assumption,illumina,hiseq_era,full_length,poly_a,unknown,sc,single_cell_plate,smartseq,,United States,2017-11-02,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
44158,SRR6250292,SRX3357214,SRS2656296,SRP123526,PRJNA416939,Single cell RNAseq SMART seq2 of wild type TLAB and MZoep tz57 zebrafish embryos at 50% epiboly stage,GSE106466,Transcriptome Analysis,SMART seq2 was performed on single cells isolated from visually staged zebrafish embryos. Overall design: Samples were all sequenced in one batch. Some were generated with a five prime' UMI tagged method  and others are full length SMART seq2.,parent bioproject:PRJNA417291,pubmed:29700225,,oep p3 S74,GSM2838378,,tissue:Embryonic cell|developmental stage:50% epiboly|library preparation batch:p3|genotype:MZoep tz57,oep p3 S74,Bases 1 8 the UMI and template switch were trimmed from the remaining reads  trimmed stretches of polyA or polyT longer than 6bp and removed any reads that were shorter than 15 bp post trimming. We identified and depleted mtRNA and rRNA reads from the remaining reads by using Bowtie2 to align all reads against a transcriptome of mtRNA and rRNA  using the parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed   no unal. Remaining reads were then aligned to a modified version of the Ensembl Zv10 release 81 zebrafish transcriptome using Bowtie2 with parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed. The transcriptome start site of each gene was padded by 100bp to increase alignment rates  since reads were generated from the 5’ end of the transcript  and not all transcripts had correct annotations for their 5’ end. Additionally  pseudogenes were removed from the transcriptome as they sometimes interfered with real genes  given our 5’ coordinate selection described below. We chose the best alignment per read as the read with the most 5’ alignment coordinate to a transcript among those that aligned to the correct strand and had an alignment score of no more than 4 below the best alignment score overall. We then iteratively collapsed pairs of UMIs  if the sum of the minimum Phred quality scores across non matching bases was at most 30 to account for mis matched bases that were likely due to sequencing errors. A collapsed pair was represented for future comparisons by the UMI with a larger prior number of aligned reads unless the other UMI had at least 75% as many reads aligned and had equal or better UMI quality scores for those alignments evaluated as the softmin over aligned reads of the softmax over quality scores for the UMI bases in an aligned read. post collapsing  the number of transcripts observed per gene was calculated. This value was then adjusted based on the probability of a collision the drawing at random of the same UMI multiple times  which was calculated from the number of UMIs not detected for a given gene. The full length fragmented reads without xxx’ enrichment or UMI were aligned using RSEM against a reference transcriptome  which is depleted of mtRNA  rRNA  and pseudogenes as described above. The command ‘rsem calculate expression’ was used with parameters  p 4   paired end   seed length 24 for performing the alignment. The resulting gene levels in FPKM was used to build the expression matrix. Genome build: GRCz10 dr81 Supplementary files format and content: Tab delimited text files include gene levels UMI counts or FKPM for each sample.,Embryonic cell,,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for 5’ enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,Fish of each genotype were incrossed and embryos were collected. Embryos were dechorionated and allowed to grow at 28°C in standard fish water containing methylene blue till 50% epiboly stage.,developmental stage:50% epiboly|library preparation batch:p3|genotype:MZoep tz57,GSM2838378,GSM2838378: oep p3 S74; Danio rerio; RNA Seq,GSM2838378,,1,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for five prime enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina HiSeq 2500,,SRP123526,,,oep_p3_S74.R1.fastq.gz oep_p3_S74.R2.fastq.gz,fastq fastq,59001376.0,1053596.0,GSM2838378 r1,0:24 1:32,A:15341220;C:13918217;G:14205278;T:15489081;N:47580,24,32,,,15341220,13918217,14205278,15489081,47580,SRX3357214,SRS2656296,SRA627905,GEO,"Schier Lab, Molecular and Cellular Biology, Harvard University",2,0.78652,0.8396,0.16905,0.18338,0.86151,0.86145,0.72891,0.73043,24,32,B,B,biological fallback assumption,illumina,hiseq_era,full_length,poly_a,unknown,sc,single_cell_plate,smartseq,,United States,2017-11-02,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
44159,SRR6250291,SRX3357213,SRS2656295,SRP123526,PRJNA416939,Single cell RNAseq SMART seq2 of wild type TLAB and MZoep tz57 zebrafish embryos at 50% epiboly stage,GSE106466,Transcriptome Analysis,SMART seq2 was performed on single cells isolated from visually staged zebrafish embryos. Overall design: Samples were all sequenced in one batch. Some were generated with a five prime' UMI tagged method  and others are full length SMART seq2.,parent bioproject:PRJNA417291,pubmed:29700225,,oep p3 S73,GSM2838377,,tissue:Embryonic cell|developmental stage:50% epiboly|library preparation batch:p3|genotype:MZoep tz57,oep p3 S73,Bases 1 8 the UMI and template switch were trimmed from the remaining reads  trimmed stretches of polyA or polyT longer than 6bp and removed any reads that were shorter than 15 bp post trimming. We identified and depleted mtRNA and rRNA reads from the remaining reads by using Bowtie2 to align all reads against a transcriptome of mtRNA and rRNA  using the parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed   no unal. Remaining reads were then aligned to a modified version of the Ensembl Zv10 release 81 zebrafish transcriptome using Bowtie2 with parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed. The transcriptome start site of each gene was padded by 100bp to increase alignment rates  since reads were generated from the 5’ end of the transcript  and not all transcripts had correct annotations for their 5’ end. Additionally  pseudogenes were removed from the transcriptome as they sometimes interfered with real genes  given our 5’ coordinate selection described below. We chose the best alignment per read as the read with the most 5’ alignment coordinate to a transcript among those that aligned to the correct strand and had an alignment score of no more than 4 below the best alignment score overall. We then iteratively collapsed pairs of UMIs  if the sum of the minimum Phred quality scores across non matching bases was at most 30 to account for mis matched bases that were likely due to sequencing errors. A collapsed pair was represented for future comparisons by the UMI with a larger prior number of aligned reads unless the other UMI had at least 75% as many reads aligned and had equal or better UMI quality scores for those alignments evaluated as the softmin over aligned reads of the softmax over quality scores for the UMI bases in an aligned read. post collapsing  the number of transcripts observed per gene was calculated. This value was then adjusted based on the probability of a collision the drawing at random of the same UMI multiple times  which was calculated from the number of UMIs not detected for a given gene. The full length fragmented reads without xxx’ enrichment or UMI were aligned using RSEM against a reference transcriptome  which is depleted of mtRNA  rRNA  and pseudogenes as described above. The command ‘rsem calculate expression’ was used with parameters  p 4   paired end   seed length 24 for performing the alignment. The resulting gene levels in FPKM was used to build the expression matrix. Genome build: GRCz10 dr81 Supplementary files format and content: Tab delimited text files include gene levels UMI counts or FKPM for each sample.,Embryonic cell,,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for 5’ enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,Fish of each genotype were incrossed and embryos were collected. Embryos were dechorionated and allowed to grow at 28°C in standard fish water containing methylene blue till 50% epiboly stage.,developmental stage:50% epiboly|library preparation batch:p3|genotype:MZoep tz57,GSM2838377,GSM2838377: oep p3 S73; Danio rerio; RNA Seq,GSM2838377,,1,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for five prime enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina HiSeq 2500,,SRP123526,,,oep_p3_S73.R1.fastq.gz oep_p3_S73.R2.fastq.gz,fastq fastq,69861456.0,1247526.0,GSM2838377 r1,0:24 1:32,A:17763729;C:16822790;G:17127718;T:18091717;N:55502,24,32,,,17763729,16822790,17127718,18091717,55502,SRX3357213,SRS2656295,SRA627905,GEO,"Schier Lab, Molecular and Cellular Biology, Harvard University",2,0.78514,0.84006,0.19824,0.21258,0.84709,0.84571,0.66962,0.67035,24,32,B,B,biological fallback assumption,illumina,hiseq_era,full_length,poly_a,unknown,sc,single_cell_plate,smartseq,,United States,2017-11-02,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
44160,SRR6250290,SRX3357212,SRS2656294,SRP123526,PRJNA416939,Single cell RNAseq SMART seq2 of wild type TLAB and MZoep tz57 zebrafish embryos at 50% epiboly stage,GSE106466,Transcriptome Analysis,SMART seq2 was performed on single cells isolated from visually staged zebrafish embryos. Overall design: Samples were all sequenced in one batch. Some were generated with a five prime' UMI tagged method  and others are full length SMART seq2.,parent bioproject:PRJNA417291,pubmed:29700225,,oep p3 S72,GSM2838376,,tissue:Embryonic cell|developmental stage:50% epiboly|library preparation batch:p3|genotype:MZoep tz57,oep p3 S72,Bases 1 8 the UMI and template switch were trimmed from the remaining reads  trimmed stretches of polyA or polyT longer than 6bp and removed any reads that were shorter than 15 bp post trimming. We identified and depleted mtRNA and rRNA reads from the remaining reads by using Bowtie2 to align all reads against a transcriptome of mtRNA and rRNA  using the parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed   no unal. Remaining reads were then aligned to a modified version of the Ensembl Zv10 release 81 zebrafish transcriptome using Bowtie2 with parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed. The transcriptome start site of each gene was padded by 100bp to increase alignment rates  since reads were generated from the 5’ end of the transcript  and not all transcripts had correct annotations for their 5’ end. Additionally  pseudogenes were removed from the transcriptome as they sometimes interfered with real genes  given our 5’ coordinate selection described below. We chose the best alignment per read as the read with the most 5’ alignment coordinate to a transcript among those that aligned to the correct strand and had an alignment score of no more than 4 below the best alignment score overall. We then iteratively collapsed pairs of UMIs  if the sum of the minimum Phred quality scores across non matching bases was at most 30 to account for mis matched bases that were likely due to sequencing errors. A collapsed pair was represented for future comparisons by the UMI with a larger prior number of aligned reads unless the other UMI had at least 75% as many reads aligned and had equal or better UMI quality scores for those alignments evaluated as the softmin over aligned reads of the softmax over quality scores for the UMI bases in an aligned read. post collapsing  the number of transcripts observed per gene was calculated. This value was then adjusted based on the probability of a collision the drawing at random of the same UMI multiple times  which was calculated from the number of UMIs not detected for a given gene. The full length fragmented reads without xxx’ enrichment or UMI were aligned using RSEM against a reference transcriptome  which is depleted of mtRNA  rRNA  and pseudogenes as described above. The command ‘rsem calculate expression’ was used with parameters  p 4   paired end   seed length 24 for performing the alignment. The resulting gene levels in FPKM was used to build the expression matrix. Genome build: GRCz10 dr81 Supplementary files format and content: Tab delimited text files include gene levels UMI counts or FKPM for each sample.,Embryonic cell,,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for 5’ enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,Fish of each genotype were incrossed and embryos were collected. Embryos were dechorionated and allowed to grow at 28°C in standard fish water containing methylene blue till 50% epiboly stage.,developmental stage:50% epiboly|library preparation batch:p3|genotype:MZoep tz57,GSM2838376,GSM2838376: oep p3 S72; Danio rerio; RNA Seq,GSM2838376,,1,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for five prime enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina HiSeq 2500,,SRP123526,,,oep_p3_S72.R1.fastq.gz oep_p3_S72.R2.fastq.gz,fastq fastq,76297816.0,1362461.0,GSM2838376 r1,0:24 1:32,A:19487447;C:18321600;G:18646618;T:19779503;N:62648,24,32,,,19487447,18321600,18646618,19779503,62648,SRX3357212,SRS2656294,SRA627905,GEO,"Schier Lab, Molecular and Cellular Biology, Harvard University",2,0.79547,0.85005,0.18506,0.20204,0.83966,0.84011,0.6719,0.67451,24,32,B,B,biological fallback assumption,illumina,hiseq_era,full_length,poly_a,unknown,sc,single_cell_plate,smartseq,,United States,2017-11-02,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
44161,SRR6250289,SRX3357211,SRS2656293,SRP123526,PRJNA416939,Single cell RNAseq SMART seq2 of wild type TLAB and MZoep tz57 zebrafish embryos at 50% epiboly stage,GSE106466,Transcriptome Analysis,SMART seq2 was performed on single cells isolated from visually staged zebrafish embryos. Overall design: Samples were all sequenced in one batch. Some were generated with a five prime' UMI tagged method  and others are full length SMART seq2.,parent bioproject:PRJNA417291,pubmed:29700225,,oep p3 S71,GSM2838375,,tissue:Embryonic cell|developmental stage:50% epiboly|library preparation batch:p3|genotype:MZoep tz57,oep p3 S71,Bases 1 8 the UMI and template switch were trimmed from the remaining reads  trimmed stretches of polyA or polyT longer than 6bp and removed any reads that were shorter than 15 bp post trimming. We identified and depleted mtRNA and rRNA reads from the remaining reads by using Bowtie2 to align all reads against a transcriptome of mtRNA and rRNA  using the parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed   no unal. Remaining reads were then aligned to a modified version of the Ensembl Zv10 release 81 zebrafish transcriptome using Bowtie2 with parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed. The transcriptome start site of each gene was padded by 100bp to increase alignment rates  since reads were generated from the 5’ end of the transcript  and not all transcripts had correct annotations for their 5’ end. Additionally  pseudogenes were removed from the transcriptome as they sometimes interfered with real genes  given our 5’ coordinate selection described below. We chose the best alignment per read as the read with the most 5’ alignment coordinate to a transcript among those that aligned to the correct strand and had an alignment score of no more than 4 below the best alignment score overall. We then iteratively collapsed pairs of UMIs  if the sum of the minimum Phred quality scores across non matching bases was at most 30 to account for mis matched bases that were likely due to sequencing errors. A collapsed pair was represented for future comparisons by the UMI with a larger prior number of aligned reads unless the other UMI had at least 75% as many reads aligned and had equal or better UMI quality scores for those alignments evaluated as the softmin over aligned reads of the softmax over quality scores for the UMI bases in an aligned read. post collapsing  the number of transcripts observed per gene was calculated. This value was then adjusted based on the probability of a collision the drawing at random of the same UMI multiple times  which was calculated from the number of UMIs not detected for a given gene. The full length fragmented reads without xxx’ enrichment or UMI were aligned using RSEM against a reference transcriptome  which is depleted of mtRNA  rRNA  and pseudogenes as described above. The command ‘rsem calculate expression’ was used with parameters  p 4   paired end   seed length 24 for performing the alignment. The resulting gene levels in FPKM was used to build the expression matrix. Genome build: GRCz10 dr81 Supplementary files format and content: Tab delimited text files include gene levels UMI counts or FKPM for each sample.,Embryonic cell,,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for 5’ enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,Fish of each genotype were incrossed and embryos were collected. Embryos were dechorionated and allowed to grow at 28°C in standard fish water containing methylene blue till 50% epiboly stage.,developmental stage:50% epiboly|library preparation batch:p3|genotype:MZoep tz57,GSM2838375,GSM2838375: oep p3 S71; Danio rerio; RNA Seq,GSM2838375,,1,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for five prime enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina HiSeq 2500,,SRP123526,,,oep_p3_S71.R1.fastq.gz oep_p3_S71.R2.fastq.gz,fastq fastq,75927544.0,1355849.0,GSM2838375 r1,0:24 1:32,A:20268836;C:17477159;G:17830475;T:20289895;N:61179,24,32,,,20268836,17477159,17830475,20289895,61179,SRX3357211,SRS2656293,SRA627905,GEO,"Schier Lab, Molecular and Cellular Biology, Harvard University",2,0.7933,0.84253,0.14163,0.15232,0.86342,0.86405,0.68068,0.68035,24,32,B,B,biological fallback assumption,illumina,hiseq_era,full_length,poly_a,unknown,sc,single_cell_plate,smartseq,,United States,2017-11-02,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
44162,SRR6250288,SRX3357210,SRS2656292,SRP123526,PRJNA416939,Single cell RNAseq SMART seq2 of wild type TLAB and MZoep tz57 zebrafish embryos at 50% epiboly stage,GSE106466,Transcriptome Analysis,SMART seq2 was performed on single cells isolated from visually staged zebrafish embryos. Overall design: Samples were all sequenced in one batch. Some were generated with a five prime' UMI tagged method  and others are full length SMART seq2.,parent bioproject:PRJNA417291,pubmed:29700225,,oep p3 S70,GSM2838374,,tissue:Embryonic cell|developmental stage:50% epiboly|library preparation batch:p3|genotype:MZoep tz57,oep p3 S70,Bases 1 8 the UMI and template switch were trimmed from the remaining reads  trimmed stretches of polyA or polyT longer than 6bp and removed any reads that were shorter than 15 bp post trimming. We identified and depleted mtRNA and rRNA reads from the remaining reads by using Bowtie2 to align all reads against a transcriptome of mtRNA and rRNA  using the parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed   no unal. Remaining reads were then aligned to a modified version of the Ensembl Zv10 release 81 zebrafish transcriptome using Bowtie2 with parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed. The transcriptome start site of each gene was padded by 100bp to increase alignment rates  since reads were generated from the 5’ end of the transcript  and not all transcripts had correct annotations for their 5’ end. Additionally  pseudogenes were removed from the transcriptome as they sometimes interfered with real genes  given our 5’ coordinate selection described below. We chose the best alignment per read as the read with the most 5’ alignment coordinate to a transcript among those that aligned to the correct strand and had an alignment score of no more than 4 below the best alignment score overall. We then iteratively collapsed pairs of UMIs  if the sum of the minimum Phred quality scores across non matching bases was at most 30 to account for mis matched bases that were likely due to sequencing errors. A collapsed pair was represented for future comparisons by the UMI with a larger prior number of aligned reads unless the other UMI had at least 75% as many reads aligned and had equal or better UMI quality scores for those alignments evaluated as the softmin over aligned reads of the softmax over quality scores for the UMI bases in an aligned read. post collapsing  the number of transcripts observed per gene was calculated. This value was then adjusted based on the probability of a collision the drawing at random of the same UMI multiple times  which was calculated from the number of UMIs not detected for a given gene. The full length fragmented reads without xxx’ enrichment or UMI were aligned using RSEM against a reference transcriptome  which is depleted of mtRNA  rRNA  and pseudogenes as described above. The command ‘rsem calculate expression’ was used with parameters  p 4   paired end   seed length 24 for performing the alignment. The resulting gene levels in FPKM was used to build the expression matrix. Genome build: GRCz10 dr81 Supplementary files format and content: Tab delimited text files include gene levels UMI counts or FKPM for each sample.,Embryonic cell,,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for 5’ enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,Fish of each genotype were incrossed and embryos were collected. Embryos were dechorionated and allowed to grow at 28°C in standard fish water containing methylene blue till 50% epiboly stage.,developmental stage:50% epiboly|library preparation batch:p3|genotype:MZoep tz57,GSM2838374,GSM2838374: oep p3 S70; Danio rerio; RNA Seq,GSM2838374,,1,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for five prime enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina HiSeq 2500,,SRP123526,,,oep_p3_S70.R1.fastq.gz oep_p3_S70.R2.fastq.gz,fastq fastq,84525840.0,1509390.0,GSM2838374 r1,0:24 1:32,A:22708087;C:19303312;G:19687748;T:22760729;N:65964,24,32,,,22708087,19303312,19687748,22760729,65964,SRX3357210,SRS2656292,SRA627905,GEO,"Schier Lab, Molecular and Cellular Biology, Harvard University",2,0.79226,0.84226,0.14243,0.15278,0.85297,0.85342,0.69822,0.70056,24,32,B,B,biological fallback assumption,illumina,hiseq_era,full_length,poly_a,unknown,sc,single_cell_plate,smartseq,,United States,2017-11-02,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
44163,SRR6250287,SRX3357209,SRS2656291,SRP123526,PRJNA416939,Single cell RNAseq SMART seq2 of wild type TLAB and MZoep tz57 zebrafish embryos at 50% epiboly stage,GSE106466,Transcriptome Analysis,SMART seq2 was performed on single cells isolated from visually staged zebrafish embryos. Overall design: Samples were all sequenced in one batch. Some were generated with a five prime' UMI tagged method  and others are full length SMART seq2.,parent bioproject:PRJNA417291,pubmed:29700225,,oep p3 S6,GSM2838373,,tissue:Embryonic cell|developmental stage:50% epiboly|library preparation batch:p3|genotype:MZoep tz57,oep p3 S6,Bases 1 8 the UMI and template switch were trimmed from the remaining reads  trimmed stretches of polyA or polyT longer than 6bp and removed any reads that were shorter than 15 bp post trimming. We identified and depleted mtRNA and rRNA reads from the remaining reads by using Bowtie2 to align all reads against a transcriptome of mtRNA and rRNA  using the parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed   no unal. Remaining reads were then aligned to a modified version of the Ensembl Zv10 release 81 zebrafish transcriptome using Bowtie2 with parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed. The transcriptome start site of each gene was padded by 100bp to increase alignment rates  since reads were generated from the 5’ end of the transcript  and not all transcripts had correct annotations for their 5’ end. Additionally  pseudogenes were removed from the transcriptome as they sometimes interfered with real genes  given our 5’ coordinate selection described below. We chose the best alignment per read as the read with the most 5’ alignment coordinate to a transcript among those that aligned to the correct strand and had an alignment score of no more than 4 below the best alignment score overall. We then iteratively collapsed pairs of UMIs  if the sum of the minimum Phred quality scores across non matching bases was at most 30 to account for mis matched bases that were likely due to sequencing errors. A collapsed pair was represented for future comparisons by the UMI with a larger prior number of aligned reads unless the other UMI had at least 75% as many reads aligned and had equal or better UMI quality scores for those alignments evaluated as the softmin over aligned reads of the softmax over quality scores for the UMI bases in an aligned read. post collapsing  the number of transcripts observed per gene was calculated. This value was then adjusted based on the probability of a collision the drawing at random of the same UMI multiple times  which was calculated from the number of UMIs not detected for a given gene. The full length fragmented reads without xxx’ enrichment or UMI were aligned using RSEM against a reference transcriptome  which is depleted of mtRNA  rRNA  and pseudogenes as described above. The command ‘rsem calculate expression’ was used with parameters  p 4   paired end   seed length 24 for performing the alignment. The resulting gene levels in FPKM was used to build the expression matrix. Genome build: GRCz10 dr81 Supplementary files format and content: Tab delimited text files include gene levels UMI counts or FKPM for each sample.,Embryonic cell,,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for 5’ enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,Fish of each genotype were incrossed and embryos were collected. Embryos were dechorionated and allowed to grow at 28°C in standard fish water containing methylene blue till 50% epiboly stage.,developmental stage:50% epiboly|library preparation batch:p3|genotype:MZoep tz57,GSM2838373,GSM2838373: oep p3 S6; Danio rerio; RNA Seq,GSM2838373,,1,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for five prime enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina HiSeq 2500,,SRP123526,,,oep_p3_S6.R1.fastq.gz oep_p3_S6.R2.fastq.gz,fastq fastq,64042216.0,1143611.0,GSM2838373 r1,0:24 1:32,A:17477673;C:14427234;G:14658115;T:17428140;N:51054,24,32,,,17477673,14427234,14658115,17428140,51054,SRX3357209,SRS2656291,SRA627905,GEO,"Schier Lab, Molecular and Cellular Biology, Harvard University",2,0.80076,0.84841,0.12096,0.12822,0.85318,0.85334,0.7865,0.78575,24,32,B,B,biological fallback assumption,illumina,hiseq_era,full_length,poly_a,unknown,sc,single_cell_plate,smartseq,,United States,2017-11-02,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
44164,SRR6250286,SRX3357208,SRS2656290,SRP123526,PRJNA416939,Single cell RNAseq SMART seq2 of wild type TLAB and MZoep tz57 zebrafish embryos at 50% epiboly stage,GSE106466,Transcriptome Analysis,SMART seq2 was performed on single cells isolated from visually staged zebrafish embryos. Overall design: Samples were all sequenced in one batch. Some were generated with a five prime' UMI tagged method  and others are full length SMART seq2.,parent bioproject:PRJNA417291,pubmed:29700225,,oep p3 S69,GSM2838372,,tissue:Embryonic cell|developmental stage:50% epiboly|library preparation batch:p3|genotype:MZoep tz57,oep p3 S69,Bases 1 8 the UMI and template switch were trimmed from the remaining reads  trimmed stretches of polyA or polyT longer than 6bp and removed any reads that were shorter than 15 bp post trimming. We identified and depleted mtRNA and rRNA reads from the remaining reads by using Bowtie2 to align all reads against a transcriptome of mtRNA and rRNA  using the parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed   no unal. Remaining reads were then aligned to a modified version of the Ensembl Zv10 release 81 zebrafish transcriptome using Bowtie2 with parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed. The transcriptome start site of each gene was padded by 100bp to increase alignment rates  since reads were generated from the 5’ end of the transcript  and not all transcripts had correct annotations for their 5’ end. Additionally  pseudogenes were removed from the transcriptome as they sometimes interfered with real genes  given our 5’ coordinate selection described below. We chose the best alignment per read as the read with the most 5’ alignment coordinate to a transcript among those that aligned to the correct strand and had an alignment score of no more than 4 below the best alignment score overall. We then iteratively collapsed pairs of UMIs  if the sum of the minimum Phred quality scores across non matching bases was at most 30 to account for mis matched bases that were likely due to sequencing errors. A collapsed pair was represented for future comparisons by the UMI with a larger prior number of aligned reads unless the other UMI had at least 75% as many reads aligned and had equal or better UMI quality scores for those alignments evaluated as the softmin over aligned reads of the softmax over quality scores for the UMI bases in an aligned read. post collapsing  the number of transcripts observed per gene was calculated. This value was then adjusted based on the probability of a collision the drawing at random of the same UMI multiple times  which was calculated from the number of UMIs not detected for a given gene. The full length fragmented reads without xxx’ enrichment or UMI were aligned using RSEM against a reference transcriptome  which is depleted of mtRNA  rRNA  and pseudogenes as described above. The command ‘rsem calculate expression’ was used with parameters  p 4   paired end   seed length 24 for performing the alignment. The resulting gene levels in FPKM was used to build the expression matrix. Genome build: GRCz10 dr81 Supplementary files format and content: Tab delimited text files include gene levels UMI counts or FKPM for each sample.,Embryonic cell,,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for 5’ enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,Fish of each genotype were incrossed and embryos were collected. Embryos were dechorionated and allowed to grow at 28°C in standard fish water containing methylene blue till 50% epiboly stage.,developmental stage:50% epiboly|library preparation batch:p3|genotype:MZoep tz57,GSM2838372,GSM2838372: oep p3 S69; Danio rerio; RNA Seq,GSM2838372,,1,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for five prime enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina HiSeq 2500,,SRP123526,,,oep_p3_S69.R1.fastq.gz oep_p3_S69.R2.fastq.gz,fastq fastq,6585488.0,117598.0,GSM2838372 r1,0:24 1:32,A:1730253;C:1555031;G:1562489;T:1731898;N:5817,24,32,,,1730253,1555031,1562489,1731898,5817,SRX3357208,SRS2656290,SRA627905,GEO,"Schier Lab, Molecular and Cellular Biology, Harvard University",2,0.75286,0.82078,0.15462,0.17297,0.91445,0.9139,0.77583,0.77999,24,32,B,B,biological fallback assumption,illumina,hiseq_era,full_length,poly_a,unknown,sc,single_cell_plate,smartseq,,United States,2017-11-02,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
44165,SRR6250285,SRX3357207,SRS2656289,SRP123526,PRJNA416939,Single cell RNAseq SMART seq2 of wild type TLAB and MZoep tz57 zebrafish embryos at 50% epiboly stage,GSE106466,Transcriptome Analysis,SMART seq2 was performed on single cells isolated from visually staged zebrafish embryos. Overall design: Samples were all sequenced in one batch. Some were generated with a five prime' UMI tagged method  and others are full length SMART seq2.,parent bioproject:PRJNA417291,pubmed:29700225,,oep p3 S68,GSM2838371,,tissue:Embryonic cell|developmental stage:50% epiboly|library preparation batch:p3|genotype:MZoep tz57,oep p3 S68,Bases 1 8 the UMI and template switch were trimmed from the remaining reads  trimmed stretches of polyA or polyT longer than 6bp and removed any reads that were shorter than 15 bp post trimming. We identified and depleted mtRNA and rRNA reads from the remaining reads by using Bowtie2 to align all reads against a transcriptome of mtRNA and rRNA  using the parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed   no unal. Remaining reads were then aligned to a modified version of the Ensembl Zv10 release 81 zebrafish transcriptome using Bowtie2 with parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed. The transcriptome start site of each gene was padded by 100bp to increase alignment rates  since reads were generated from the 5’ end of the transcript  and not all transcripts had correct annotations for their 5’ end. Additionally  pseudogenes were removed from the transcriptome as they sometimes interfered with real genes  given our 5’ coordinate selection described below. We chose the best alignment per read as the read with the most 5’ alignment coordinate to a transcript among those that aligned to the correct strand and had an alignment score of no more than 4 below the best alignment score overall. We then iteratively collapsed pairs of UMIs  if the sum of the minimum Phred quality scores across non matching bases was at most 30 to account for mis matched bases that were likely due to sequencing errors. A collapsed pair was represented for future comparisons by the UMI with a larger prior number of aligned reads unless the other UMI had at least 75% as many reads aligned and had equal or better UMI quality scores for those alignments evaluated as the softmin over aligned reads of the softmax over quality scores for the UMI bases in an aligned read. post collapsing  the number of transcripts observed per gene was calculated. This value was then adjusted based on the probability of a collision the drawing at random of the same UMI multiple times  which was calculated from the number of UMIs not detected for a given gene. The full length fragmented reads without xxx’ enrichment or UMI were aligned using RSEM against a reference transcriptome  which is depleted of mtRNA  rRNA  and pseudogenes as described above. The command ‘rsem calculate expression’ was used with parameters  p 4   paired end   seed length 24 for performing the alignment. The resulting gene levels in FPKM was used to build the expression matrix. Genome build: GRCz10 dr81 Supplementary files format and content: Tab delimited text files include gene levels UMI counts or FKPM for each sample.,Embryonic cell,,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for 5’ enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,Fish of each genotype were incrossed and embryos were collected. Embryos were dechorionated and allowed to grow at 28°C in standard fish water containing methylene blue till 50% epiboly stage.,developmental stage:50% epiboly|library preparation batch:p3|genotype:MZoep tz57,GSM2838371,GSM2838371: oep p3 S68; Danio rerio; RNA Seq,GSM2838371,,1,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for five prime enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina HiSeq 2500,,SRP123526,,,oep_p3_S68.R1.fastq.gz oep_p3_S68.R2.fastq.gz,fastq fastq,51989560.0,928385.0,GSM2838371 r1,0:24 1:32,A:13525099;C:12235022;G:12510508;T:13677564;N:41367,24,32,,,13525099,12235022,12510508,13677564,41367,SRX3357207,SRS2656289,SRA627905,GEO,"Schier Lab, Molecular and Cellular Biology, Harvard University",2,0.79086,0.84275,0.1571,0.16871,0.88185,0.88252,0.67648,0.67296,24,32,B,B,biological fallback assumption,illumina,hiseq_era,full_length,poly_a,unknown,sc,single_cell_plate,smartseq,,United States,2017-11-02,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
44166,SRR6250284,SRX3357206,SRS2656288,SRP123526,PRJNA416939,Single cell RNAseq SMART seq2 of wild type TLAB and MZoep tz57 zebrafish embryos at 50% epiboly stage,GSE106466,Transcriptome Analysis,SMART seq2 was performed on single cells isolated from visually staged zebrafish embryos. Overall design: Samples were all sequenced in one batch. Some were generated with a five prime' UMI tagged method  and others are full length SMART seq2.,parent bioproject:PRJNA417291,pubmed:29700225,,oep p3 S67,GSM2838370,,tissue:Embryonic cell|developmental stage:50% epiboly|library preparation batch:p3|genotype:MZoep tz57,oep p3 S67,Bases 1 8 the UMI and template switch were trimmed from the remaining reads  trimmed stretches of polyA or polyT longer than 6bp and removed any reads that were shorter than 15 bp post trimming. We identified and depleted mtRNA and rRNA reads from the remaining reads by using Bowtie2 to align all reads against a transcriptome of mtRNA and rRNA  using the parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed   no unal. Remaining reads were then aligned to a modified version of the Ensembl Zv10 release 81 zebrafish transcriptome using Bowtie2 with parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed. The transcriptome start site of each gene was padded by 100bp to increase alignment rates  since reads were generated from the 5’ end of the transcript  and not all transcripts had correct annotations for their 5’ end. Additionally  pseudogenes were removed from the transcriptome as they sometimes interfered with real genes  given our 5’ coordinate selection described below. We chose the best alignment per read as the read with the most 5’ alignment coordinate to a transcript among those that aligned to the correct strand and had an alignment score of no more than 4 below the best alignment score overall. We then iteratively collapsed pairs of UMIs  if the sum of the minimum Phred quality scores across non matching bases was at most 30 to account for mis matched bases that were likely due to sequencing errors. A collapsed pair was represented for future comparisons by the UMI with a larger prior number of aligned reads unless the other UMI had at least 75% as many reads aligned and had equal or better UMI quality scores for those alignments evaluated as the softmin over aligned reads of the softmax over quality scores for the UMI bases in an aligned read. post collapsing  the number of transcripts observed per gene was calculated. This value was then adjusted based on the probability of a collision the drawing at random of the same UMI multiple times  which was calculated from the number of UMIs not detected for a given gene. The full length fragmented reads without xxx’ enrichment or UMI were aligned using RSEM against a reference transcriptome  which is depleted of mtRNA  rRNA  and pseudogenes as described above. The command ‘rsem calculate expression’ was used with parameters  p 4   paired end   seed length 24 for performing the alignment. The resulting gene levels in FPKM was used to build the expression matrix. Genome build: GRCz10 dr81 Supplementary files format and content: Tab delimited text files include gene levels UMI counts or FKPM for each sample.,Embryonic cell,,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for 5’ enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,Fish of each genotype were incrossed and embryos were collected. Embryos were dechorionated and allowed to grow at 28°C in standard fish water containing methylene blue till 50% epiboly stage.,developmental stage:50% epiboly|library preparation batch:p3|genotype:MZoep tz57,GSM2838370,GSM2838370: oep p3 S67; Danio rerio; RNA Seq,GSM2838370,,1,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for five prime enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina HiSeq 2500,,SRP123526,,,oep_p3_S67.R1.fastq.gz oep_p3_S67.R2.fastq.gz,fastq fastq,84747488.0,1513348.0,GSM2838370 r1,0:24 1:32,A:22025430;C:20007980;G:20351968;T:22292024;N:70086,24,32,,,22025430,20007980,20351968,22292024,70086,SRX3357206,SRS2656288,SRA627905,GEO,"Schier Lab, Molecular and Cellular Biology, Harvard University",2,0.79678,0.84797,0.17303,0.1864,0.85551,0.85525,0.6823,0.67888,24,32,B,B,biological fallback assumption,illumina,hiseq_era,full_length,poly_a,unknown,sc,single_cell_plate,smartseq,,United States,2017-11-02,Gastrula,Embryo,Embryo Imprecise,All anatomical structures
44167,SRR6250283,SRX3357205,SRS2656287,SRP123526,PRJNA416939,Single cell RNAseq SMART seq2 of wild type TLAB and MZoep tz57 zebrafish embryos at 50% epiboly stage,GSE106466,Transcriptome Analysis,SMART seq2 was performed on single cells isolated from visually staged zebrafish embryos. Overall design: Samples were all sequenced in one batch. Some were generated with a five prime' UMI tagged method  and others are full length SMART seq2.,parent bioproject:PRJNA417291,pubmed:29700225,,oep p3 S66,GSM2838369,,tissue:Embryonic cell|developmental stage:50% epiboly|library preparation batch:p3|genotype:MZoep tz57,oep p3 S66,Bases 1 8 the UMI and template switch were trimmed from the remaining reads  trimmed stretches of polyA or polyT longer than 6bp and removed any reads that were shorter than 15 bp post trimming. We identified and depleted mtRNA and rRNA reads from the remaining reads by using Bowtie2 to align all reads against a transcriptome of mtRNA and rRNA  using the parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed   no unal. Remaining reads were then aligned to a modified version of the Ensembl Zv10 release 81 zebrafish transcriptome using Bowtie2 with parameters  q   phred33  N 1  I 1  X 2000  k 200   score min L 3.36  0.617   no discordant   no mixed. The transcriptome start site of each gene was padded by 100bp to increase alignment rates  since reads were generated from the 5’ end of the transcript  and not all transcripts had correct annotations for their 5’ end. Additionally  pseudogenes were removed from the transcriptome as they sometimes interfered with real genes  given our 5’ coordinate selection described below. We chose the best alignment per read as the read with the most 5’ alignment coordinate to a transcript among those that aligned to the correct strand and had an alignment score of no more than 4 below the best alignment score overall. We then iteratively collapsed pairs of UMIs  if the sum of the minimum Phred quality scores across non matching bases was at most 30 to account for mis matched bases that were likely due to sequencing errors. A collapsed pair was represented for future comparisons by the UMI with a larger prior number of aligned reads unless the other UMI had at least 75% as many reads aligned and had equal or better UMI quality scores for those alignments evaluated as the softmin over aligned reads of the softmax over quality scores for the UMI bases in an aligned read. post collapsing  the number of transcripts observed per gene was calculated. This value was then adjusted based on the probability of a collision the drawing at random of the same UMI multiple times  which was calculated from the number of UMIs not detected for a given gene. The full length fragmented reads without xxx’ enrichment or UMI were aligned using RSEM against a reference transcriptome  which is depleted of mtRNA  rRNA  and pseudogenes as described above. The command ‘rsem calculate expression’ was used with parameters  p 4   paired end   seed length 24 for performing the alignment. The resulting gene levels in FPKM was used to build the expression matrix. Genome build: GRCz10 dr81 Supplementary files format and content: Tab delimited text files include gene levels UMI counts or FKPM for each sample.,Embryonic cell,,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for 5’ enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,Fish of each genotype were incrossed and embryos were collected. Embryos were dechorionated and allowed to grow at 28°C in standard fish water containing methylene blue till 50% epiboly stage.,developmental stage:50% epiboly|library preparation batch:p3|genotype:MZoep tz57,GSM2838369,GSM2838369: oep p3 S66; Danio rerio; RNA Seq,GSM2838369,,1,50% epiboly embryos were manually deyolked and mechanically dissociated. Single cells were manually picked under a dissecting microscope and flash frozen on dry ice in 4uL lysis buffer. Libraries were prepared using a SMART seq2 protocol Picelli et al.  2014 with custom TSO oligos and indexing adapters Satija et al.  2015. Reaction volumn for tagmentation steps were reduced to 15% of the volumn indicated in the protocol. 88 MZoep libraries were built with customized indices for five prime enriched transcripts with unique molecular identifiers UMIs. The remaining libraries were built without xxx  and contained fragments along full length transcripts.,,RNA-Seq,TRANSCRIPTOMIC,cDNA,PAIRED,ILLUMINA,Illumina HiSeq 2500,,SRP123526,,,oep_p3_S66.R1.fastq.gz oep_p3_S66.R2.fastq.gz,fastq fastq,74525864.0,1330819.0,GSM2838369 r1,0:24 1:32,A:19200379;C:17757638;G:18093711;T:19414533;N:59603,24,32,,,19200379,17757638,18093711,19414533,59603,SRX3357205,SRS2656287,SRA627905,GEO,"Schier Lab, Molecular and Cellular Biology, Harvard University",2,0.79564,0.84753,0.18356,0.1978,0.85614,0.85518,0.68007,0.67596,24,32,B,B,biological fallback assumption,illumina,hiseq_era,full_length,poly_a,unknown,sc,single_cell_plate,smartseq,,United States,2017-11-02,Gastrula,Embryo,Embryo Imprecise,All anatomical structures