Hi Heidi, Thanks for setting me straight - I in fact have a panel of 96 unrepeated features, used to produce experiments with a variable number of replicates per group. This works very well. Thanks! On 6/26/12 12:14 PM, "Heidi Dvinge" <heidi at="" ebi.ac.uk=""> wrote: >Hi Kipper, > >> Dear List, >> >> I am using HTqPCR to analyze Fluidigm data (specifically, the 96*96 >> format). I have been using specifically limmaCtData to look for >> differential expression between groups, always having the same number of >> replicates. However, I currently have a dataset of several groups with >>3 >> replicates and 1 group with 2 replicates. As far as I can tell, >> limmaCtData expects the number of replicates to be the same between >>groups >> (i.e. there is only one 'ndups' parameter). Has anyone else encountered >> this issue? >> >ndups actually refers to the number of replicate *features*, i.e. genes, >on your plate. As long as you use the same platform for all your samples, >this should be constant. > >The number of samples within each group is indicated using the design >matrix in limmaCtData, and the parameter 'groups' in ttestCtData. So in >limmaCtData you can just use a design matrix along the lines of: > >> samples <- rep(c("treat1", "treat2", "treat3", "control"), c(3,2,3,3)) >> design <- model.matrix(~0+samples) > >You then indicate your comparisons of interest in the contrast matrix, and >use both matrices in your call to limmaCtData. > >If I've misunderstood you and you actually have a variable number of >replicates of each gene, then I'm afraid limmaCtData isn't capable of >handling this with ndups. > >HTH >\Heidi > >> Thank you, >> Kipper Fletez-Brant >> >> _______________________________________________ >> Bioconductor mailing list >> Bioconductor at r-project.org >> https://stat.ethz.ch/mailman/listinfo/bioconductor >> Search the archives: >> http://news.gmane.org/gmane.science.biology.informatics.conductor >> > >