>Date: Thu, 9 Sep 2004 22:01:22 +1000 (EST) >From: "Gordon K Smyth" <smyth@wehi.edu.au> >Subject: Re: [BioC] Finding and controlling for dye bias in Limma >To: "Sean Davis" <sdavis2@mail.nih.gov> >Cc: Bioconductor <bioconductor@stat.math.ethz.ch> > >> We have several samples hybed on two-color arrays (treated versus >> untreated). Half are in dye-swap. Can we use Limma to estimate the >> effect of dye-bias? > >Yes > >> A design matrix might look something like: >> >> Array Treatment Dye-bias >> 1 1 0 >> 2 1 0 >> 3 1 0 >> 4 1 0 >> 5 -1 1 >> 6 -1 1 >> 7 -1 1 >> 8 -1 1 >> >> The second coefficient would be significant for "dye-biased genes" >> while the first would be treatment effect while controlling for dye >> bias? Do I understand correctly? > >No, the Dye-bias column should be all 1's. If he has half the samples in each dye-swap status, would it be alternating 1's and -1's? Array Treatment Dye-bias 1 1 1 2 1 -1 3 1 1 4 1 -1 5 -1 1 6 -1 -1 7 -1 1 8 -1 -1 >Gordon > >> Thanks, >> Sean