I have to analyse some data from MG_U74 chips that contain 2613 bad probesets (antisense DNA strand) resulted from a mistake of Affymetrix. I need to remove these bad probeset from CEL file before I regenerate data (from CEL files ) because some algorithms e.g. RMA, may use these bad probesets to do normalization. I called Affymetrix but was told that their mask files can remove bad probesets only after data is regerated from CEL files. For anybody who has experience in such a thing or has knowledge on how to resolve this issue, please give some suggestions. Hairong [[alternative HTML version deleted]]

At least as far as rma() goes, you can use the "subset" argument to specify a list of probesets to restrict to for all calculations (ie the other probesets and their probe intensities are entirely ignored). Ben On Mon, 2005-03-14 at 15:52 -0600, Hairong Wei wrote: > I have to analyse some data from MG_U74 chips that contain 2613 bad > probesets (antisense DNA strand) resulted from a mistake of Affymetrix. > I need to remove these bad probeset from CEL file before I regenerate data > (from CEL files ) because some algorithms e.g. RMA, may use these bad > probesets to do normalization. I called Affymetrix but was told that > their mask files can remove bad probesets only after data is regerated from > CEL files. > > For anybody who has experience in such a thing or has knowledge on how to > resolve this issue, please give some suggestions. > > Hairong > > > > > > > [[alternative HTML version deleted]] > > _______________________________________________ > Bioconductor mailing list > Bioconductor@stat.math.ethz.ch > https://stat.ethz.ch/mailman/listinfo/bioconductor

Hi, Ben: Thank you very much for help. Your suggestion for using 'subset' to remove bad probesets works very well for RMA(). For anyone who knows how to run dChip or mas5 using a subset of good probesets on the chips, please do not be hesitated to help. I am processing MGU74A on which there are 2613 antisense probes. Hairong -----Original Message----- From: Ben Bolstad [mailto:bolstad@stat.berkeley.edu] Sent: Monday, March 14, 2005 5:00 PM To: Hairong Wei Cc: 'bioconductor@stat.math.ethz.ch' Subject: Re: [BioC] Removing bad probesets At least as far as rma() goes, you can use the "subset" argument to specify a list of probesets to restrict to for all calculations (ie the other probesets and their probe intensities are entirely ignored). Ben On Mon, 2005-03-14 at 15:52 -0600, Hairong Wei wrote: > I have to analyse some data from MG_U74 chips that contain 2613 bad > probesets (antisense DNA strand) resulted from a mistake of Affymetrix. > I need to remove these bad probeset from CEL file before I regenerate > data (from CEL files ) because some algorithms e.g. RMA, may use these bad > probesets to do normalization. I called Affymetrix but was told that > their mask files can remove bad probesets only after data is regerated from > CEL files. > > For anybody who has experience in such a thing or has knowledge on how to > resolve this issue, please give some suggestions. > > Hairong > > > > > > > [[alternative HTML version deleted]] > > _______________________________________________ > Bioconductor mailing list > Bioconductor@stat.math.ethz.ch > https://stat.ethz.ch/mailman/listinfo/bioconductor