Using Frozen RMA to normalize microarray CEL files for GPL570

normalize.R

# source("http://bioconductor.org/biocLite.R")
library("affy")
library("frma")
library("hgu133plus2frmavecs")
library("parallel")

# String -> String
change.file.extension <- function(filename, extension){
  # Strip file extensions from filename, 
  #  and replace with given extension
  base <- strsplit(filename, ".", fixed = T)[[1]][1]
  full.name = paste(base, extension, sep = ".")
  return(full.name)
}

# String -> SideEffect(FileSystem)
parse.and.save <- function(cel.filename){
  affy_data <- ReadAffy(filenames = cel.filename)
  frma_data <- frma(affy_data)
  exprs <- assayData(frma_data)$exprs
  
  save.filename = change.file.extension(cel.filename, 'csv')
  write.table(
    exprs, 
    file = save.filename, 
    quote = F, 
    row.names = T, 
    col.names = F)
}

files = list.celfiles()

mclapply(files, parse.and.save, mc.cores = 8)

normalize2.R

# Given a folder, normalize all CEL files and and save to same folder as frma.csv

folder = "/home/data/frma_test"

library("affy")
library("frma")
library("hgu133plus2frmavecs")

# String -> SideEffect(FileSystem)
parse.and.save <- function(cel.filenames){
  affy_data <- ReadAffy(filenames = cel.filenames)
  frma_data <- frma(affy_data)
  exprs <- assayData(frma_data)$exprs
  save.filename = paste(folder, "/frma.csv", sep='')
  write.csv(
    exprs,
    file = save.filename
}
    
files <- list.celfiles(folder, full.names = T)

parse.and.save(files)

mclapply (multi-core left apply) is an R function to map a function over a list in parallel.