GSE14333 Processing Pipeline

Publication

Life science alliance (2020) — PMID 32817263

Dataset

Expression data from 290 primary colorectal cancers

1. 1

   Use the simpleaffy package in R/Bioconductor to calculate MAS5.0 calls.

   R v4.3 (Bioconductor 3.18)

   $ Bash example

   # Install R and Bioconductor simpleaffy package if not already installed
   # For Conda (recommended for environment management):
   # conda create -n bioconductor_env r-base bioconductor-simpleaffy -y
   # conda activate bioconductor_env
   
   # For R directly:
   # Rscript -e 'if (!requireNamespace("BiocManager", quietly = TRUE)) install.packages("BiocManager")'
   # Rscript -e 'BiocManager::install("simpleaffy")'
   
   # Create an R script to perform MAS5.0 normalization
   cat << 'EOF' > mas5_normalization.R
   library(simpleaffy)
   
   # Define input directory containing .CEL files and output file
   # Assuming .CEL files are in the current directory. Adjust 'cel_dir' if needed.
   cel_dir <- "."
   output_file <- "mas5_normalized_expression.tsv"
   
   # List all .CEL files in the specified directory
   cel_files <- list.celfiles(cel_dir, full.names=TRUE)
   
   # Check if any .CEL files were found
   if (length(cel_files) == 0) {
     stop(paste("No .CEL files found in", cel_dir, ". Please ensure .CEL files are present or adjust 'cel_dir'."))
   }
   
   # Read Affymetrix .CEL files into an AffyBatch object
   # For more complex experiments (e.g., with sample metadata), consider creating a phenoData file
   # and using read.affybatch(filenames=cel_files, phenoData=pheno_data_object)
   raw_data <- read.affybatch(filenames=cel_files)
   
   # Perform MAS5.0 normalization
   # By default, MAS5 normalization in simpleaffy outputs log2 transformed values.
   mas5_normalized_data <- mas5(raw_data)
   
   # Extract expression values
   expression_matrix <- exprs(mas5_normalized_data)
   
   # Write normalized expression matrix to a TSV file
   write.table(expression_matrix, file=output_file, sep="\t", quote=FALSE, row.names=TRUE)
   
   message(paste("MAS5.0 normalized expression values written to:", output_file))
   EOF
   
   # Execute the R script
   Rscript mas5_normalization.R

   Copy
2. 2

   These values were subsequently normalized using quantile normalization.

   R (with preprocessCore package) (Inferred with models/gemini-2.5-flash) vR 4.3.0, preprocessCore 1.62.0 GitHub

   $ Bash example

   # Install R and preprocessCore if not already available
   # conda create -n r_env r-base bioconductor-preprocesscore -y
   # conda activate r_env
   
   # Example: Create a dummy input file (replace with your actual input_matrix.tsv)
   # echo -e "gene\tsample1\tsample2\tsample3" > input_matrix.tsv
   # echo -e "geneA\t100\t200\t50" >> input_matrix.tsv
   # echo -e "geneB\t50\t100\t25" >> input_matrix.tsv
   # echo -e "geneC\t200\t50\t100" >> input_matrix.tsv
   
   # R script for quantile normalization
   Rscript -e '
     library(preprocessCore)
     
     # Define input and output file names
     input_file <- "input_matrix.tsv"
     output_file <- "normalized_matrix.tsv"
     
     # Read the input matrix
     # Assuming the first column contains row identifiers (e.g., gene names)
     # and subsequent columns contain numeric data for samples.
     # header=TRUE assumes the first row contains column names (sample IDs).
     # sep="\t" assumes tab-separated values.
     # Adjust these parameters based on the actual input file format.
     input_data <- read.table(input_file, sep="\t", header=TRUE, row.names=1, check.names=FALSE)
     
     # Convert to matrix for preprocessCore
     data_matrix <- as.matrix(input_data)
     
     # Perform quantile normalization
     normalized_matrix <- normalize.quantiles(data_matrix)
     
     # Restore column names (sample IDs) and row names (gene IDs)
     colnames(normalized_matrix) <- colnames(data_matrix)
     rownames(normalized_matrix) <- rownames(data_matrix)
     
     # Write the normalized matrix to an output file
     # quote=FALSE prevents R from adding quotes around string values.
     # col.names=NA is used when row.names=TRUE to leave the top-left cell empty,
     # which is standard for matrices with row names written to file.
     write.table(normalized_matrix, output_file, sep="\t", quote=FALSE, row.names=TRUE, col.names=NA)
   '

   Copy View on GitHub

Tools Used