Hi, I have a RNAseq dataset with this design: condition treatment sample1 A treat sample2 A treat sample3 A con sample4 A con sample5 B treat sample6 B treat sample7 B con sample8 B con so there are two conditions and within each condition treatment and control. So far I simply compared treat vs. con within each group, so I get logFCs and p values for A and B separately (for this I actually subsetted the dataset and used ~treatment as design) Now I would also like compare the two conditions to see genes that respond differently to treatment in A and in B. So my question: how do I set up the design for that? Thanks in advance! max

hi Max, To compare the response to treatment in A and B you would use an interaction term: design(dds) <- ~ condition + treatment + condition:treatment DESeq(dds) results(dds) # this last line is all that is needed because the interaction term is last in the design formula # equivalently, you can check the name in resultsNames(dds) and then it will be something like results(dds, name="conditionB.treatmenttreat") Small p-values for the interaction term indicate that the log fold change due to treatment is significantly different for the two conditions. Mike On Thu, Mar 6, 2014 at 6:38 AM, Kauer Max <maximilian.kauer@ccri.at> wrote: > Hi, > I have a RNAseq dataset with this design: > > condition treatment > sample1 A treat > sample2 A treat > sample3 A con > sample4 A con > sample5 B treat > sample6 B treat > sample7 B con > sample8 B con > > > so there are two conditions and within each condition treatment and > control. > So far I simply compared treat vs. con within each group, so I get logFCs > and > p values for A and B separately (for this I actually subsetted the dataset > and used ~treatment as design) > > Now I would also like compare the two conditions to see genes that respond > differently to treatment in A and in B. > So my question: how do I set up the design for that? > > Thanks in advance! > max > > _______________________________________________ > Bioconductor mailing list > Bioconductor@r-project.org > https://stat.ethz.ch/mailman/listinfo/bioconductor > Search the archives: > http://news.gmane.org/gmane.science.biology.informatics.conductor > [[alternative HTML version deleted]]