Only recently I have posted the problem of not having the corresponding Affy IDS and Locus IDS to the significant nodes from a GOHyperG application. I think this is a problem every body encounters sooner or later when working with gene ontologies. I have a temporary solution which performs quite well. So here is the code: input: GOLOCmap: This is the result from applying GOALLLOCUSID to the set of significant GO IDS. Its a list and every element corresponds to a certain GO ID and the entries to the LOCUS IDS mapping on it. LOC: This are the AffyIDS and the corresponding LOCUS IDS, a vector. It comes from applying hgu133aLOCUSID to a set of AffyIDS. The function looks for the Locus IDS (LOC) in the entries of every element of the list GOLOCmap. The result is a list. Every element corresponds to a certain GO ID and the entries are the ocurring AffyIDS. Locus_Affy_fromGO<-function(GOLOCmap,LOC) { LOCs<-sort(LOC) maximum<-max(LOCs) output<-as.list(1:length(names(GOLOCmap))) res<-0 for (i in 1:length(names(GOLOCmap))) { output[[i]]<-list(GOID=names(GOLOCmap)[[i]],LOCUSIDSAffyIDS=c(1,2)) GOLOCmaps<-sort(GOLOCmap[[i]]) j<-1 while(GOLOCmaps[j]<= maximum) { if(j>=length(GOLOCmaps)) break res<-c(res,LOCs[LOCs==GOLOCmaps[j]]) j<-j+1 } output[[i]]$LOCUSIDSAffyIDS <-res[-1] res<-0 } output } I hope it is useful for some of you. And please let me know if any bugs are occuring. Best regards Michael