Hi all, I was pretty sure this was already out there on the mailing list, but i can't seem to find the post.. I get a reproducible error when calling sam() from siggenes (latest siggenes v1.2.17) with RMA-normalized data. It throws the error on 133aV2 chips with only 2 reps (not with 3) and with hu95 chips with 3 replicates. The error msg is: >Error in "names<-.default('*tmp*', value = c("100_g_at", "1000_at", >"1001_at", : 'names' attribute [12625] must be the same length as the >vector [12551] I haven't dug thru the siggenes code yet b/c I think this error has been discussed here - doesn't this happen when multiple replicate samples have identical expression values (which can happen with RMA, due to the median polish) ? Then with SAM, using Wilcoxon you get some probes with identical ranks and when the ranks are sorted, some are equal so they drop out, but the names vector is still the same length, so it crashes...? Just wondering if that's what is happening... Thanks for any help, Ken

Hi Ken, you are right: There is a bug in the computation of the Wilcoxon rank sums when a gene has to be removed because of zero variance or too many NAs (sorry for my last mail to you in which I've said that there is no bug). I will fix this bug today. So it should be available as soon as the devel version for BioC 1.8 is available. Please let me know if you need this package immediately. Best, Holger > --- Urspr?ngliche Nachricht --- > Von: "Ken Termiso" <jerk_alert at="" hotmail.com=""> > An: bioconductor at stat.math.ethz.ch > Betreff: [BioC] siggenes error with rma & wilcoxon test > Datum: Fri, 14 Oct 2005 17:04:32 +0000 > > Hi all, > > I was pretty sure this was already out there on the mailing list, but i > can't seem to find the post.. > > I get a reproducible error when calling sam() from siggenes (latest > siggenes > v1.2.17) with RMA-normalized data. It throws the error on 133aV2 chips > with > only 2 reps (not with 3) and with hu95 chips with 3 replicates. > > The error msg is: > > >Error in "names<-.default('*tmp*', value = c("100_g_at", "1000_at", > >"1001_at", : 'names' attribute [12625] must be the same length as the > >vector [12551] > > I haven't dug thru the siggenes code yet b/c I think this error has been > discussed here - doesn't this happen when multiple replicate samples have > identical expression values (which can happen with RMA, due to the median > polish) ? Then with SAM, using Wilcoxon you get some probes with > identical > ranks and when the ranks are sorted, some are equal so they drop out, but > the names vector is still the same length, so it crashes...? > > Just wondering if that's what is happening... > > Thanks for any help, > Ken > > _______________________________________________ > Bioconductor mailing list > Bioconductor at stat.math.ethz.ch > https://stat.ethz.ch/mailman/listinfo/bioconductor > --