Hello, this sounds like a rather trivial question but somehow the an answer eludes me. I am analysing gene expression data with limma in a two-factor model with interactions. No problems there. What I would like to do is to first cluster the genes (using my own clustering procedure that I'm working on), and then search for clusters that are differentially expressed (e.g. factor A is significant in cluster X, or interaction is significant in cluster Y etc.). I have concocted my own bootstrapping procedure, but would prefer to use an established tool if possible. I remember that there was a tool or package for randomisation-based GO analysis in R which allowed to specify arbitrary linear models for comparisons, but can't seem to be able to find it right now. On the other hand, I could treat the assignments to clusters like assignments to GO categories and just test the significant genes for enrichment. Which approach and which package would you recommend? j. -- -------- January Weiner -------------------------------------- [[alternative HTML version deleted]]

The function you are trying to remember may be ?roast Gordon > Date: Thu, 7 Nov 2013 14:13:38 +0100 > From: January Weiner <january.weiner at="" gmail.com=""> > To: Bioconductor mailing list <bioconductor at="" r-project.org=""> > Subject: [BioC] Differential gene expression of a cluster or group of > genes > > Hello, > > this sounds like a rather trivial question but somehow the an answer > eludes me. I am analysing gene expression data with limma in a > two-factor model with interactions. No problems there. > > What I would like to do is to first cluster the genes (using my own > clustering procedure that I'm working on), and then search for clusters > that are differentially expressed (e.g. factor A is significant in > cluster X, or interaction is significant in cluster Y etc.). I have > concocted my own bootstrapping procedure, but would prefer to use an > established tool if possible. > > I remember that there was a tool or package for randomisation-based GO > analysis in R which allowed to specify arbitrary linear models for > comparisons, but can't seem to be able to find it right now. > > On the other hand, I could treat the assignments to clusters like > assignments to GO categories and just test the significant genes for > enrichment. > > Which approach and which package would you recommend? > > j. > > -------- January Weiner -------------------------------------- ______________________________________________________________________ The information in this email is confidential and intend...{{dropped:4}}