Merging the results of a statistical test in different conditions and calculating p-value

Question

Suppose we have two mouse strains S1 and S2 and we have already tested the effect of two different conditions (say treatment vs. control) on the level of gene expression of any strain, each experiment with several replicates.

Now we are looking for the genes that are for example significantly higher expressed in treatment condition in both strains. I have already selected the genes that are higher expressed in the treatment condition for each of the strains, and calculated the related p-value.

A simple idea is to find the intersection - the genes that are higher expressed in both strains. Is this a good method? How to compute a p-value? The answers that can be easily computed in R are more than welcome!

Answer 1

Maybe this reference Statistical tests for the intersection of independent lists of genes: Sensitivity, FDR, and type I error control and others in the same issue of this journal may be helpful.