Hi, I have a very customerized GSEA (gene set enrichment analysis) project and want to try hypergeometric distr as one of my approaches. After "carefully" reading documents and related discussions from this list, I think the following assignments are right and hope it to be confirmed before I put into codes: N: the universe, or the "urn", which is the number of genes in the intersection between "union of genes in all GO terms" AND "your chip". n: the number of the selected x genes (most of time, differentially expressed genes), reduced by N (which means removing all genes not in N) D: success events for a specific GO term in universe, i.e. all genes in a specific GO term, again reduced by removing all genes not in N. k: success events in the selected genes for a specific GO term, reduced by removing all genes not in N. I think the key part here is, everything defined in a general hypergeometric distr needs to be reduced by chip. Am I right? Please correct me if I am wrong. Thanks. -- Weiwei Shi, Ph.D Research Scientist GeneGO, Inc. "Did you always know?" "No, I did not. But I believed..." ---Matrix III