I' m planning to test vsn normalization on affymetrix arrays with RMA background correction. But from the vsn vignette and other messages in this list it seems to me that vsn on affymetrix chips is used without background correction: e=expresso(data,bg.correct=FALSE,normalize.method="vsn",pmcorrect.meth od ="pmonly",summary.method="medianpolish") There is a reason for this? Is it wrong to use expresso in this way: e=expresso(data,bg.correct="rma",normalize.method="vsn",pmcorrect.meth od ="pmonly",summary.method="medianpolish") ? Thank you Remo

in principle it does not sound like a good idea since vsn does something that could be considered bg correction. in any case i have already seen the result of this and you get very similar results to just using RMA (quantile normalization). look at the RMAVSN entry in http://affycomp.biostat.jhsph.edu On Jul 19, 2004, at 3:56 PM, Remo Sanges wrote: > I' m planning to test vsn normalization on affymetrix arrays > with RMA background correction. > > But from the vsn vignette and other messages in this list > it seems to me that vsn on affymetrix chips is used without > background correction: > e=expresso(data,bg.correct=FALSE,normalize.method="vsn",pmcorrect.me tho > d="pmonly",summary.method="medianpolish") > > There is a reason for this? > > Is it wrong to use expresso in this way: > e=expresso(data,bg.correct="rma",normalize.method="vsn",pmcorrect.me tho > d="pmonly",summary.method="medianpolish") ? > > Thank you > > Remo > > _______________________________________________ > Bioconductor mailing list > Bioconductor@stat.math.ethz.ch > https://www.stat.math.ethz.ch/mailman/listinfo/bioconductor