Hi all, I have a question about making contrasts using limma package. I want to look at the differences of 2 age groups in male and female respectively, so I have an interaction term in my model matrix, I am not sure whether my contrast matrix is correct or not (I don't have Female_AgeGroup1 and Male_AgeGroup1 columns in my design matrix), so not sure how to make contrast matrix for this analysis. The R codes were attached as follows, thank you so much!

design <- model.matrix(~ 0+FetalSex+FetalSex:AgeGroups+ArrayDateBatch)

colnames(design) <- c("Female", "Male", "Batch2", "Batch3", "Female_AgeGroup2", "Male_AgeGroup2")

contrastMatrix <- makeContrasts(M_F = Male-Female, F_G2_G1 = Female_AgeGroup2 - (Female-Female_AgeGroup2), M_G2_G1 = Male_AgeGroup2 - (Male-Male_AgeGroup2), levels = design)

You would be much better off to generate four groups, one for each sex/age group and then make whatever contrasts you want, rather than trying to figure out what your existing coefficients are.