I am using justRMA to analyze Affy rat2302 chips. I do have installed on my system the alternative MBNI cdf package but do not want to use it in this case. For some reason, just.rma and justRMA are using the MBNI cdf as the default cdf file and using the parameter "cdfname = 'rat2302'" does not work to over-ride. Below is my code and sessionInfo. Note that I am using the devel versions or R and BioC with all latest package updates installed. Thanks, Mark > affy.object.preprocessed$eSet <-just.rma(filenames=list.celfiles(), phenoData = read.pData.ordered.func("pdata.3.txt")) Background correcting Normalizing Calculating Expression > affy.object.preprocessed$eSet Expression Set (exprSet) with 10255 genes 30 samples phenoData object with 2 variables and 30 cases varLabels Condition: read from file Treatment: read from file > annotation(affy.object.preprocessed$eSet) [1] "rn230rnentrezg6" > sessionInfo() Version 2.3.0 Under development (unstable) (2006-01-01 r36947) i386-pc-mingw32 attached base packages: [1] "splines" "tools" "methods" "stats" "graphics" "grDevices" "utils" "datasets" "base" other attached packages: rn230rnentrezg6cdf affycoretools GOstats multtest genefilter survival xtable RBGL "6.0.0" "1.3.1" "1.5.4" "1.8.0" "1.9.2" "2.20" "1.3-0" "1.7.7" annotate GO graph Ruuid cluster limma affy Biobase "1.9.2" "1.6.5" "1.9.9" "1.9.0" "1.10.2" "2.4.9" "1.9.7" "1.9.6" > Mark W. Kimpel MD

Hi Mark, Kimpel, Mark William wrote: > I am using justRMA to analyze Affy rat2302 chips. I do have installed on > my system the alternative MBNI cdf package but do not want to use it in > this case. For some reason, just.rma and justRMA are using the MBNI cdf > as the default cdf file and using the parameter "cdfname = 'rat2302'" > does not work to over-ride. Below is my code and sessionInfo. Note that > I am using the devel versions or R and BioC with all latest package > updates installed. > > Thanks, Mark > > >>affy.object.preprocessed$eSet <-just.rma(filenames=list.celfiles(), > > phenoData = read.pData.ordered.func("pdata.3.txt")) This is sort of an unusual thing to do. Why are you putting your exprSet into a list? Also, what does read.pData.ordered.func() do? I can't find this function. Anyway, I can't reproduce what you are seeing: > eset <- just.rma(filenames=list.celfiles()) Background correcting Normalizing Calculating Expression Warning message: Incompatible phenoData object. Created a new one. in: just.rma(filenames = list.celfiles()) > annotation(eset) [1] "mouse4302" > eset <- just.rma(filenames=list.celfiles(), cdfname="mm430mm3enst6") Attaching package: 'mm430mm3enst6cdf' The following object(s) are masked from package:mouse4302cdf : i2xy xy2i Background correcting Normalizing Calculating Expression Warning message: Incompatible phenoData object. Created a new one. in: just.rma(filenames = list.celfiles(), cdfname = "mm430mm3enst6") > annotation(eset) [1] "mm430mm3enst6" > eset <- just.rma(filenames=list.celfiles(), cdfname="mouse4302") Background correcting Normalizing Calculating Expression Warning message: Incompatible phenoData object. Created a new one. in: just.rma(filenames = list.celfiles(), cdfname = "mouse4302") > annotation(eset) [1] "mouse4302" > sessionInfo() Version 2.3.0 Under development (unstable) (2006-01-01 r36947) i386-pc-mingw32 attached base packages: [1] "splines" "tools" "methods" "stats" "graphics" "grDevices" "utils" [8] "datasets" "base" other attached packages: mm430mm3enst6cdf mouse4302cdf affycoretools GOstats multtest "6.0.0" "1.10.0" "1.3.1" "1.5.4" "1.9.3" genefilter survival xtable RBGL annotate "1.9.1" "2.20" "1.3-0" "1.7.5" "1.9.2" GO graph Ruuid cluster limma "1.11.0" "1.9.4" "1.9.0" "1.10.2" "2.4.7" affy Biobase "1.9.7" "1.9.4" Best, Jim -- James W. MacDonald Affymetrix and cDNA Microarray Core University of Michigan Cancer Center 1500 E. Medical Center Drive 7410 CCGC Ann Arbor MI 48109 734-647-5623

Jim, The problem appears to be a corrupt data file as I cannot reproduce the error with other datasets. To answer your other questions... I put the eset into a list with another slot reserved for annotation information. That way I only have to build the annotation once and then store it in a database with the eset. Maybe a bit odd, but I'm not a trained programmer and it works for me. The read.pData.ordered.func function is something I wrote to make sure that the proper phenotype information was correctly assigned to samples. Not long ago I needed to sort my pdata file by phenotype to check something and forgot to resort it by sampleName before saving it. After getting some puzzling analysis results I discovered that read.phenoData does not match sampleNames with CEL file names (in my case one and the same) so I wrote this function to do that. I will share the code, simple as it is, in case this is beneficial to anyone else (see below). Thanks, Mark read.pData.ordered.func<-function(pdata.file) { pD <- read.phenoData(filename = pdata.file, sep = "\t", header=TRUE, row.names=1) cel.files<-list.celfiles() for (i in 1:length(cel.files)) { cel.files[i]<-substr(x=cel.files[i], start=1, stop=(nchar(cel.files[i])-4)) } pos<-match(cel.files, rownames(pD at pData)) pD<-pD[pos] pD }