Hello, I am working with a set of Affymetrix arrays, all of the same type, among which there is for each sample a biological duplicate. I was wondering when I should "group" the duplicates : before normalizing ? After ? I can't either find how to do it properly, do you have suggestions ? Thanks a lot for the help you may provide ! Amelie -- The Wellcome Trust Sanger Institute is operated by Genome Research Limited, a charity registered in England with number 1021457 and a company registered in England with number 2742969, whose registered office is 215 Euston Road, London, NW1 2BE.

On Fri, Mar 27, 2009 at 6:04 AM, <ab19@sanger.ac.uk> wrote: > Hello, > > I am working with a set of Affymetrix arrays, all of the same type, among > which there is for each sample a biological duplicate. I was wondering > when I should "group" the duplicates : before normalizing ? After ? I > can't either find how to do it properly, do you have suggestions ? > > Thanks a lot for the help you may provide ! > Hi, Amelie. You'll want to deal with the replicates only when you are dealing with differential expression. See, for example, the limma user guide which has some examples of interest. Sean [[alternative HTML version deleted]]

On Fri, Mar 27, 2009 at 7:07 AM, <ab19@sanger.ac.uk> wrote: > Dear Sean, > > Thank you for your answer. I understand that you need replicates when you > work with differential expression in order to assess the the variance > between them and compare it then with the "inter" variance. My purpose is > not to perform a differential expression analysis. Still, is it good to > loose the information that replicates may provide but using only one of > them ? > Generally, you'll want to use all the samples that meet quality criteria. And it is probably not useful to do things like "averaging" replicates and the like. If you need more details, then you might explain what you are trying to accomplish. Sean > > Thanks for your help ! > > Amelie > > > On Fri, Mar 27, 2009 at 6:04 AM, <ab19@sanger.ac.uk> wrote: > > > >> Hello, > >> > >> I am working with a set of Affymetrix arrays, all of the same type, > >> among > >> which there is for each sample a biological duplicate. I was wondering > >> when I should "group" the duplicates : before normalizing ? After ? I > >> can't either find how to do it properly, do you have suggestions ? > >> > >> Thanks a lot for the help you may provide ! > >> > > > > Hi, Amelie. > > > > You'll want to deal with the replicates only when you are dealing with > > differential expression. See, for example, the limma user guide which > has > > some examples of interest. > > > > Sean > > > > > > > -- > The Wellcome Trust Sanger Institute is operated by Genome Research > Limited, a charity registered in England with number 1021457 and a > company registered in England with number 2742969, whose registered > office is 215 Euston Road, London, NW1 2BE. > [[alternative HTML version deleted]]

Hi Amelie: You can use gcRMA to normalize your raw data. Your duplicates can be "grouped" when you perform differential expression analysis using packages such as Limma which will fit linear models for each gene. Hope this helps. Cheers, Wei ab19 at sanger.ac.uk wrote: > Hello, > > I am working with a set of Affymetrix arrays, all of the same type, among > which there is for each sample a biological duplicate. I was wondering > when I should "group" the duplicates : before normalizing ? After ? I > can't either find how to do it properly, do you have suggestions ? > > Thanks a lot for the help you may provide ! > > Amelie > > > >