Hi Adai, Thanks for getting back to me. I have about 100 HGU133A and HGUB and another 100 HGU plus2. I am sure that I can normalize each set separately, then union all of them. My concern will be the batch effects within each set. Can I do a secondary normalization using some simple approach trying to remove the batch effects among the sets? Thanks, A.G. Lee -----Original Message----- From: Adaikalavan Ramasamy [mailto:ramasamy@cancer.org.uk] Sent: Wednesday, April 12, 2006 7:00 PM To: Li, Aiguo (NIH/NCI) [C] Cc: bioconductor at stat.math.ethz.ch Subject: Re: [BioC] A questions related to affymetrix chip normalizationfrom multiple platform How many samples do you have with HGU133A, HGU133B and HGU133 plus 2.0 ? You might have to preprocess the 3 different chip types separately and then merge using union approach. This means that the samples from A and B chip type would have missing values. MAS 5.0 is outdated and has been shown to perform poorly for low-abundance genes. Use RMA, GCRMA or some other modern algorithm. Regards, Adai On Tue, 2006-04-11 at 17:57 -0400, Li, Aiguo (NIH/NCI) [C] wrote: > Hello all. > > > > I have a set of chips using affymetrix platform. Some of the chips are > from U133A and U133B and the others are plus2 chips. As I understand, > all the probesets in U133A and U133B are included in plus2 chips. The > question now is how I can normalize the cel files from 2 or 3 platforms. > I saw people have used MAS-5 output with a common scaling factor, for > example 500, for all platforms. Is there a better way than that? > > > > Thanks in advance! > > > > A.G. Lee > > > [[alternative HTML version deleted]] > > _______________________________________________ > Bioconductor mailing list > Bioconductor at stat.math.ethz.ch > https://stat.ethz.ch/mailman/listinfo/bioconductor > Search the archives: http://news.gmane.org/gmane.science.biology.informatics.conductor >