Dear Limma Experts: I am currently working on a project that has exactly the same design as the example given in Section 9.7 multi-level experiments of Limma package's user guide. But our questions about the contrasts are slightly different: 1) what's the overall difference between disease and normal regardless of tissue types, and 2) what's the overall difference between tissue A and tissue B regardless of whether diseased or normal? Here is my modification to the contrasts in attempt to answer our questions: cm <- makeContrasts( + DiseasedvsNormal = (Diseased.A-Normal.A + Diseased.B-Normal.B)/2, + TissueBvsTissueA = (Normal.B-Normal.A + Diseased.B-Diseased.A)/2, + levels = design); Am I right? Tremendous thanks for your help. Yu Cai -- output of sessionInfo(): -- Sent via the guest posting facility at bioconductor.org.

That is correct.