i think this is a pretty good article about microarrays: http://www.nature.com/cgi- taf/DynaPage.taf?file=/nrg/journal/v5/n3/full/nrg1297_fs.html this brings up a question of mine which is how to best measure the signal to noise ratio. also, how do you decide how many replicates are required for an experiment with out knowing (there is no way of knowing right?) what the variance will be? -isaac --

have you looked at: http://affycomp.biostat.jhsph.edu/ ? we have various measures of singal and various measures of precission (noise) for comparing expression measures. having just one singal to noise ratio may be a bit simplistic but you can look at the areas under the ROC curves for the task of detecting differentially expressed genes as bottom line results that depends directly on the signal to noise ratios. On Wed, 3 Mar 2004, Isaac Mehl wrote: > i think this is a pretty good article about microarrays: > http://www.nature.com/cgi- taf/DynaPage.taf?file=/nrg/journal/v5/n3/full/nrg1297_fs.html > this brings up a question of mine which is how to best measure the > signal to noise ratio. also, how do you decide how many replicates are > required for an experiment with out knowing (there is no way of knowing > right?) what the variance will be? > > -isaac > > -- > > _______________________________________________ > Bioconductor mailing list > Bioconductor@stat.math.ethz.ch > https://www.stat.math.ethz.ch/mailman/listinfo/bioconductor >