Hi everyone, Sorry to be redundant, but I'm just making sure: GCOS uses the median of three balanced Wilcoxon signed rank tests performed on (PM(1)-MM(1))-(PM(2)-MM(2)) values in order to determine the significance of differential expression between arrays. I'm under the impression that no package exists for bioconductor which can perform this analysis, or anything like it. Is that correct? If it is, would a limma analysis of one group with 1 replicate and one group with no replicates be considered sufficient (convincing, ;))? A two word response is good enough for me. Thanks everyone. Dave