Hi Biostars!

I am analyzing a set of samples from males and females in WGCNA, relating module eigengenes to case-control status separately, in males and females. I observed some associations between case-control status and some modules in males, but not in females. However, I noticed that I had more males than females to start with (and males had a higher proportion of cases, which could be skewing my results?). To test if my trait-module association in males was being driven by the excess in cases in the males cohort relative to the female cohort, I've subsetted the male cohort [so that this now has the same number of cases and controls that there were in the females cohort]. The number of modules observed in this new analysis is the same, but the number of genes attributed to each module eigengene has changed, as expected. Is there a way to identify the preservation/correlation between modules calculated in these two analyses?