Hello, Referring the limma user guide section 9.7-Multi-level experiments (pages 49-50),

This experiment involves 6 subjects, including 3 patients who have the disease and 3 normal subjects. From each subject, we have expression profiles of two tissue types, A and B.

Subjects=1,2,3,4,5,6 Condition: Diseased, Normal Tissue type: A, B

https://www.bioconductor.org/packages/devel/bioc/vignettes/limma/inst/doc/usersguide.pdf

I understood that you can test for Diseased vs Normal for TissueA or Diseased vs Normal for Tissue B (between patients comparison). and also you can test for TissueA vs TissueB for Normal or TissueA vs TissueB for Diseased (within patient comparison).

Assuming there is no interaction between Condition and Tissue,

1. Can I perform a main effect test for overall comparison of Diseased vs Normal, not specific to Tissue type?
2. Can I perform a main effect test for comparison of Tissue A vs Tissue B, not specific to Condition group?

If I write this contrast, will it test for 1 and 2 above

cm=makeContrasts(DiseasedvsNormal=(Disease.A+Diseased.B)/2 - (Normal.A+Normal.B)/2,
                                 TissueAvsTissueB= (Diseased.A+Normal.A)/2 - (Normal.B+Diseased.B)/2,
                                  levels=design)

Yusuf

Yes, you could call those contrasts to be "main effects", since they are equivalent to "main effect" tests in a suitably defined factorial model. Whether you call them "main effects" or not, it is obvious what the contrasts are testing.