[BioC] normalize.quantil, different results, affy_1.14.1 and affy_1.16.0

Markus Schmidberger schmidb at ibe.med.uni-muenchen.de
Fri Feb 15 12:11:25 CET 2008


Hello,

using "normalize.AffyBatch.quantiles" on R-2.5.0 with affy_1.14.1 and on 
R-2.6.0 with affy_1.16.0 generates different results! Is this correct?
I think there were only some changes in the package structure 
(normalize.quantile moved to preprocessCore). So there should be no 
differences in the results.

For testing I used the code below.

Best regards
Markus Schmidberger


#####
library(affy)
sessionInfo()

    affy   affyio  Biobase
"1.14.1"  "1.4.0" "1.14.0"

pathCELFiles <- "/home/schmidb/tmp/E-TABM-185"
celFile <- list.celfiles(path=pathCELFiles,full.names=TRUE)[1:5];
affyBatch <- ReadAffy(filenames=celFile);
affyBatchALT <- normalize.AffyBatch.quantiles(affyBatch, type="pmonly")

save(affyBatchALT, file="alt")

#####
library(affy)
sessionInfo()

[1] affy_1.16.0          preprocessCore_1.0.0 affyio_1.6.1
[4] Biobase_1.16.1

pathCELFiles <- "/home/schmidb/tmp/E-TABM-185"
celFile <- list.celfiles(path=pathCELFiles,full.names=TRUE)[1:5];
affyBatch <- ReadAffy(filenames=celFile);
affyBatchNEU <- normalize.AffyBatch.quantiles(affyBatch, type="pmonly")

save(affyBatchNEU, file="neu")

#####
load("alt")
load("neu")

all.equal(exprs(affyBatchALT),exprs(affyBatchNEU))
"Mean relative  difference: 0.0001107451"



More information about the Bioconductor mailing list