GSE125970 Processing Pipeline

Publication

eLife (2025) — PMID 40184163

Dataset

Single-cell transcriptome analysis reveals differential nutrient absorption functions in human intestine

1. 1

   Quality control and preprocessing of high throughput sequencing data were using SOAPnuke with parameters -n 0.1 -l 10 -A 0.25 -Q 2 -G --seqType 1

   SOAPnuke vNot specified GitHub

   $ Bash example

   # Install SOAPnuke (if not already installed)
   # conda install -c bioconda soapnuke
   
   # Create output directory
   mkdir -p filtered_data
   
   # Define input and output paths
   RAW_READS="raw_reads.fastq"
   OUTPUT_DIR="filtered_data"
   OUTPUT_PREFIX="sample_name" # SOAPnuke will append .clean.fastq to this
   
   # Run SOAPnuke for quality control and preprocessing
   SOAPnuke filter \
       -f "${RAW_READS}" \
       -o "${OUTPUT_DIR}" \
       -D "${OUTPUT_PREFIX}" \
       -n 0.1 \
       -l 10 \
       -A 0.25 \
       -Q 2 \
       -G \
       --seqType 1

   Copy View on GitHub
2. 2

   16 bp 10xTM Barcodes and 10bp UMIs were encoded at the start of Read 1 (R1).

   umi_tools (Inferred with models/gemini-2.5-flash) v1.1.2 GitHub

   $ Bash example

   # Install umi_tools if not already installed
   # conda install -c bioconda umi_tools
   
   # Placeholder for input and output files
   # Replace 'input_R1.fastq.gz' with your actual Read 1 file
   # Replace 'output_R1_umi_extracted.fastq.gz' with your desired output file name
   
   umi_tools extract \
       --bc-pattern=C{16} \
       --umi-pattern=N{10} \
       -I input_R1.fastq.gz \
       -S output_R1_umi_extracted.fastq.gz

   Copy View on GitHub
3. 3

   Reads alignment, filtering, barcode counting, and UMI counting were performed using cellranger 2.1 with default parameters

   Cell Ranger v2.1

   $ Bash example

   # Install Cell Ranger (example, adjust path as needed)
   # wget https://cf.10xgenomics.com/releases/cell-exp/cellranger-2.1.0.tar.gz
   # tar -xzf cellranger-2.1.0.tar.gz
   # export PATH=/path/to/cellranger-2.1.0:$PATH
   
   # Create output directory
   mkdir -p my_cellranger_output
   
   # Run cellranger count for reads alignment, filtering, barcode counting, and UMI counting
   # Replace 'path/to/fastqs' with the actual directory containing FASTQ files (e.g., /data/fastqs)
   # Replace 'path/to/transcriptome_reference' with the actual path to the Cell Ranger-compatible transcriptome reference (e.g., /ref/cellranger/refdata-cellranger-GRCh38-1.2.0)
   # Replace 'my_sample_name' with the actual sample name if your FASTQ files are not named according to 10x conventions
   cellranger count \
       --id=my_cellranger_output \
       --transcriptome=path/to/transcriptome_reference \
       --fastqs=path/to/fastqs \
       --sample=my_sample_name

   Copy
4. 4

   Sample normalization and scaled gene expression data were calculated by Seurat2.3.2

   Seurat v2.3.2 GitHub

   $ Bash example

   # Install Seurat (version 2.3.2 might require specific R/package versions)
   # It's recommended to use a specific Bioconda environment for older versions.
   # For example, to install R 3.4.x and Seurat 2.3.2:
   # conda create -n seurat232 r=3.4.4 r-seurat=2.3.2 -c conda-forge -c bioconda
   # conda activate seurat232
   
   # Create an R script for normalization and scaling
   cat << 'EOF' > run_seurat_normalization.R
   library(Seurat)
   
   # Load your raw count matrix (replace 'counts.tsv' with your actual file path)
   # The count matrix should have genes as rows and cells as columns.
   # Example: counts <- read.table("counts.tsv", sep="\t", header=TRUE, row.names=1)
   # For demonstration, let's create a dummy matrix
   set.seed(123)
   counts <- matrix(sample(0:100, 1000, replace = TRUE), ncol = 10)
   rownames(counts) <- paste0("gene", 1:100)
   colnames(counts) <- paste0("cell", 1:10)
   
   # Create a Seurat object
   # For Seurat v2, the raw.data slot was used for the initial count matrix
   seurat_obj <- CreateSeuratObject(raw.data = counts)
   
   # Normalize the data
   # LogNormalize is a common method for scRNA-seq data
   seurat_obj <- NormalizeData(object = seurat_obj, normalization.method = "LogNormalize", scale.factor = 10000)
   
   # Scale the data
   # This scales and centers the data across cells for each gene
   seurat_obj <- ScaleData(object = seurat_obj)
   
   # Save the processed Seurat object
   saveRDS(seurat_obj, "normalized_scaled_seurat_object.rds")
   
   # Optionally, print a summary or head of the scaled data
   # print(head(seurat_obj@scale.data[1:5, 1:5]))
   EOF
   
   # Execute the R script
   Rscript run_seurat_normalization.R

   Copy View on GitHub