Entering edit mode
Ben Ganzfried
▴
30
@ben-ganzfried-4902
Last seen 9.6 years ago
Hi-- apologies if this is a duplicate email. I just joined the list
today
and I didn't know how long the email I just sent before I joined would
be
under moderator approval.
As a brief introduction, I'm a research assistant in a cancer lab. I
graduated from college in 2009 and didn't see myself getting involved
in
medical research at all until I served as a bone marrow donor for my
brother
a year ago. I'm currently finishing up classes to go to medical
school and
I work in a bioinformatics lab where I'm really enjoying learning how
to use
R through fun projects.
My current project requires me to make a single object that will link
our
expression data with its associated phenotype data, and I'm a little
confused about the best approach to do this. I want to use the
pdata()
function and I have all the phenotype data available as csv files, and
I
would greatly appreciate any specific advice you may have about how to
implement this. Here is the script so far that generates expression
values,
and I want to add the single object to this ideally:
library(affy)
library(arrayQualityMetrics)
inputargs <- commandArgs(TRUE)
print(inputargs)
sessionInfo()
strInputAccession <- inputargs[1]
strBaseDir <- inputargs[2]
processedDir <-
paste(strBaseDir,strInputAccession,"PROCESSED/DEFAULT",sep="/")
qcDir <- paste(strBaseDir,strInputAccession,"QC",sep="/")
sampleFile <-
paste(processedDir,"/",strInputAccession,"_RAWfilenames.txt",sep="")
defaultExprs <-
paste(processedDir,"/",strInputAccession,"_default_exprs.csv",sep="")
if(file.exists(sampleFile)){
affyobj <- try(ReadAffy(filenames=read.table(sampleFile,as.is
=T,header=FALSE)[,1]))
if(class(affyobj)=="try-error"){
rm(affyobj)
}else{
sampleNames(affyobj) <- sub("\\..*","",sampleNames(affyobj)) #get
rid
of .CEL etc extension, just keep GSM
}
}
if(!exists("affyobj")){
mdExprs <- as.matrix(read.csv(defaultExprs,row.names=1))
if( min( mdExprs, na.rm = TRUE ) >= 0 & max( mdExprs, na.rm = TRUE )
>= 50
) {
mdExprs <- log(mdExprs, base = 2) }
affyobj <- new("ExpressionSet",exprs=mdExprs)
}
#I'd like to make one single object with phenotype and expression data
here
x <- arrayQualityMetrics(expressionset = affyobj,
outdir=qcDir,
force = TRUE,
spatial=FALSE)
[[alternative HTML version deleted]]
