I have been having problems with getting P values when using limma with affy data from the Arabidopsis AG arrays but not the ATH1 arrays. All the P values from the AG data are 0.999984 even though the t statistic is varying in value. Has anyone else had this problem? Helen