Hello, I am using RankProd to identify differentially expressed genes between mouse and human arrays, where the human genes are represented as the homologous mouse genes which are common to both platforms. Is it valid to use RankProd for such a comparison, as two classes each with single origin? It's a little unclear how classes might be handled differently from origins. From reading the vignette this application seems fine, but after reading the original paper we weren't so sure. Thank you, Adam Cornwell [[alternative HTML version deleted]]