GSE162335 Processing Pipeline

Publication

Science advances (2023) — PMID 36724232

Dataset

Transcriptional Survey of Ileal-Anal Pouch Immune Cells from Ulcerative Colitis

1. 1

   The Cellranger software suite (https://support.10xgenomics.com/single-cell-gene- expression/software/pipelines/latest/what-is-cell-ranger) from 10X was used to demultiplex cellular barcodes, align reads to the human genome (GRCh38 ensemble, http://useast.ensembl.org/Homo_sapiens/Info/Index) and perform UMI counting

   Cell Ranger vlatest

   $ Bash example

   # Install Cell Ranger (example, specific version might vary)
   # wget https://cf.10xgenomics.com/releases/cell-exp/cellranger-8.0.0.tar.gz
   # tar -xzf cellranger-8.0.0.tar.gz
   # export PATH=/path/to/cellranger-8.0.0:$PATH
   
   # Download or build the 10x Genomics reference transcriptome for GRCh38
   # The description specifies 'human genome (GRCh38 ensemble)', which corresponds to 10x Genomics' pre-built references.
   # Example for a recent GRCh38 reference:
   # wget https://cf.10xgenomics.com/releases/cell-exp/refdata-gex-GRCh38-2024-A.tar.gz
   # tar -xzf refdata-gex-GRCh38-2024-A.tar.gz
   # REF_GENOME_PATH="/path/to/refdata-gex-GRCh38-2024-A"
   
   # Placeholder for input FASTQ files directory
   FASTQ_DIR="/path/to/your/fastqs"
   # Placeholder for sample ID (e.g., the prefix of your fastq files)
   SAMPLE_ID="my_single_cell_sample"
   # Placeholder for the 10x Genomics reference transcriptome path
   # This should be a path to a directory containing the 'fasta' and 'genes' subdirectories
   REF_GENOME_PATH="/path/to/10x_genomics_refdata_gex_GRCh38_202X_A"
   
   cellranger count \
       --id="${SAMPLE_ID}_analysis" \
       --transcriptome="${REF_GENOME_PATH}" \
       --fastqs="${FASTQ_DIR}" \
       --sample="${SAMPLE_ID}" \
       --localcores=8 \
       --localmem=64

   Copy
2. 2

   From filtered counts Seurat1 version 3.1.3 was used to process the single cell data including normalization, integration, dimension reduction, UMAP representation

   UMAP v3.1.3 (via Seurat) GitHub

   $ Bash example

   # Install R if not already installed
   # sudo apt-get update && sudo apt-get install r-base
   #
   # Install Seurat and its dependencies (like uwot for UMAP) in R
   # Rscript -e 'install.packages("Seurat", repos="http://cran.us.r-project.org")'
   # Rscript -e 'install.packages("patchwork", repos="http://cran.us.r-project.org")' # Often used with Seurat
   # Rscript -e 'install.packages("uwot", repos="http://cran.us.r-project.org")' # UMAP dependency
   
   # Placeholder for filtered counts input file (e.g., a CSV or TSV matrix)
   # In a real scenario, 'filtered_counts.csv' would be provided as input.
   # Example: Create a dummy filtered_counts.csv for demonstration
   echo "gene,cell1,cell2,cell3" > filtered_counts.csv
   echo "gene1,10,20,30" >> filtered_counts.csv
   echo "gene2,5,15,25" >> filtered_counts.csv
   echo "gene3,20,10,5" >> filtered_counts.csv
   
   # R script to process single cell data using Seurat v3.1.3
   Rscript -e '
   library(Seurat)
   library(uwot) # UMAP dependency
   
   # Load filtered counts data
   # Adjust this loading based on the actual input format (e.g., 10x Genomics output, AnnData, etc.)
   # Assuming input is a CSV where the first column is gene names and subsequent columns are cell counts
   counts_df <- read.csv("filtered_counts.csv", row.names = 1)
   counts_matrix <- as.matrix(counts_df)
   
   # Create Seurat object
   seurat_obj <- CreateSeuratObject(counts = counts_matrix, project = "single_cell_analysis")
   
   # 1. Normalization
   seurat_obj <- NormalizeData(seurat_obj, normalization.method = "LogNormalize", scale.factor = 10000)
   
   # 2. Identify highly variable features
   seurat_obj <- FindVariableFeatures(seurat_obj, selection.method = "vst", nfeatures = 2000)
   
   # 3. Scale data (for PCA)
   all.genes <- rownames(seurat_obj)
   seurat_obj <- ScaleData(seurat_obj, features = all.genes)
   
   # 4. Dimension Reduction (PCA)
   seurat_obj <- RunPCA(seurat_obj, features = VariableFeatures(object = seurat_obj), npcs = 30)
   
   # 5. Integration (placeholder - typically for multiple datasets/batches)
   # The description mentions "integration". If multiple datasets were to be integrated,
   # the workflow would involve FindIntegrationAnchors and IntegrateData. For a single dataset,
   # this step might refer to batch correction if relevant metadata is available.
   # As no specific integration method or multiple datasets are mentioned, we proceed with a single-sample workflow.
   
   # 6. UMAP representation
   seurat_obj <- RunUMAP(seurat_obj, dims = 1:30) # Use first 30 PCs for UMAP
   
   # Save the processed Seurat object
   saveRDS(seurat_obj, file = "processed_seurat_object.rds")
   
   # Optional: Save UMAP coordinates to a CSV file
   umap_coords <- Embeddings(seurat_obj, reduction = "umap")
   write.csv(umap_coords, file = "umap_coordinates.csv")
   '

   Copy View on GitHub

Tools Used