{"database":"biostudies-arrayexpress","file_versions":[],"scores":null,"additional":{"submitter":["Nicolas Navarro Fraunhoffer"],"organism":["Homo sapiens"],"software":["Seurat  v5.3.0","Split-pipe pipeline v1.1.1"],"full_dataset_link":["https://www.ebi.ac.uk/biostudies/studies/E-MTAB-16856"],"description":["Single-cell transcriptomic dataset comprising 41 patient-derived cell cultures (PDCs) established from pancreatic ductal adenocarcinoma (PDAC). The dataset was generated using Evercode™ technology with the Whole Transcriptome v2 kit (Parse Biosciences, ECW02130), based on Split Pool Ligation-based Transcriptome sequencing (SPLiT-seq). For scRNA-seq, PDCs were processed in two batches, 14 samples in the first batch (R1) and 27 samples in the second batch (R2). Sequencing was performed separately for each batch using the Illumina NovaSeq X Plus platform. Paired-end reads were generated on a 10B flow cell with 200 sequencing cycles. FASTQ quality control (QC), barcode correction, alignment to the human reference genome (GRCh38.p13), multiplet detection, and raw count matrix generation were performed using the split-pipe pipeline (v1.1.1, Parse Biosciences) with default parameters. This single-cell collection of experimentally derived models reveals transcriptional heterogeneity both within and across PDCs. The submitted file consists of the count matrix along with the corresponding metadata."],"repository":["biostudies-arrayexpress"],"sample_protocol":["Library Construction - The SPLiT-seq protocol applies three rounds of combinatorial indexing to uniquely label individual cells. Briefly, for each PDC, approximately 2,000 fixed cells were used as input. Cells were first distributed into 48 wells for reverse transcription, during which the first molecular tag was incorporated. They were then pooled and redistributed twice into 96-well plates for the sequential ligation of the second and third indexes. After this indexing process, cells were evenly divided into sublibraries and lysed. The uniquely tagged cDNA molecules were isolated using a streptavidin-coated magnetic bead pulldown, followed by PCR amplification. The cDNA was then enzymatically fragmented, ligated to sequencing adapters, and assigned a sublibrary-specific fourth index during the final PCR step.","Sequencing - Sequencing was performed separately for each culture batch using the Illumina NovaSeq X Plus. Paired-end reads were generated using a 10B flow cell with 200 sequencing cycles.","Nucleic Acid Extraction - No direct RNA extraction was performed. During library construction, the cells were permeabilized, and the RNA molecules were barcoded in situ with well-specific barcode sequences.","Sample Collection - To perform scRNA-seq, PDCs were cultured in two batches, 14 samples in the first batch (R1) and 27 samples in the second batch (R2). Following thawing, each PDC was cultured for a maximum of two passages over approximately two weeks. At that point, cells at 60–70% confluency were harvested, and aliquots of 500,000 cells were pelleted, resuspended in 500 µL of freezing medium (FBS + 10% DMSO; Sigma, D8418), counted, and cryopreserved at -80°C using a Corning CoolCell, for subsequent single-cell transcriptomic analysis. Frozen samples were processed using the Evercode™ Cell Fixation v2 kit (Parse Biosciences, ECF2101) and subsequently barcoded and prepared for sequencing according to the manufacturer’s workflow provided with the Evercode™ Whole Transcriptome v2 kit (Parse Biosciences, ECW02130)."],"figure_sub":["Organization","MINSEQE Score","Assays and Data","Processed Data","MAGE-TAB Files"],"data_protocol":["Data Transformation - The raw count matrices were subsequently loaded into R using the Seurat package. To account for potential batch-specific variation, QC filtering was applied independently to each batch. Initially, duplicated barcodes were removed, retaining only one per cell. Then, for batch R1, cells with a total of expressed genes between 2,000 and 10,000 were retained. Concomitantly, for batch R2, cells with expressed genes between 1,500 and 9,000 were retained. aw counts were normalised using log normalisation with a scale factor of 10,000. The top 2,000 features were selected based on variance-stabilising transformation, and the data was subsequently scaled. Principal Component Analysis (PCA) was performed on the selected features. Harmony integration was applied using the first 37 principal components, which explained 90% of the variance in the dataset.","Sequence Alignment - FASTQ file quality control, barcode correction, alignment to the human genome (GRCh38.p13), multiplet detection, and raw count matrix generation were performed using the split-pipe pipeline with the default parameters."],"omics_type":["Unknown","Transcriptomics","Genomics","Proteomics"],"instrument_platform":["Illumina NovaSeq X","Whole Transcriptome v2 kit"],"study_type":["RNA-seq of coding RNA from single cells"],"species":["Homo sapiens"],"pubmed_authors":["Nelson Dusetti","Nicolas Navarro Fraunhoffer"],"additional_accession":[]},"is_claimable":false,"name":"Single-cell transcriptomic profiling of patient-derived pancreatic ductal adenocarcinoma primary cell cultures","description":"Single-cell transcriptomic dataset comprising 41 patient-derived cell cultures (PDCs) established from pancreatic ductal adenocarcinoma (PDAC). The dataset was generated using Evercode™ technology with the Whole Transcriptome v2 kit (Parse Biosciences, ECW02130), based on Split Pool Ligation-based Transcriptome sequencing (SPLiT-seq). For scRNA-seq, PDCs were processed in two batches, 14 samples in the first batch (R1) and 27 samples in the second batch (R2). Sequencing was performed separately for each batch using the Illumina NovaSeq X Plus platform. Paired-end reads were generated on a 10B flow cell with 200 sequencing cycles. FASTQ quality control (QC), barcode correction, alignment to the human reference genome (GRCh38.p13), multiplet detection, and raw count matrix generation were performed using the split-pipe pipeline (v1.1.1, Parse Biosciences) with default parameters. This single-cell collection of experimentally derived models reveals transcriptional heterogeneity both within and across PDCs. The submitted file consists of the count matrix along with the corresponding metadata.","dates":{"release":"2026-04-15T00:00:00Z","modification":"2026-04-17T01:01:33.887Z","creation":"2026-04-02T11:53:37.243Z"},"accession":"E-MTAB-16856","cross_references":{"ENA":["ERP191728"],"EFO":["EFO_0002944","EFO_0004170","EFO_0005684","EFO_0004917","EFO_0005518","EFO_0003816","EFO_0004184"]}}