Hi all, In previous works (1999 or 2000), Kerr and Churchill (and others) have proposed different models with fixed effects to analyze microarray data and to detect differentially expressed genes. For this, they consider the Gene effect G and the Variety effect (or treatement effect) V as fixed effects. By this way, it is possible to estimate the interaction GV... But, the Gene effect can also be considered as a random effect, and in this case the interaction GV is also a random effect and can not be estimated (only predicted with the BLUP for example). In this case, I wonder how it is possible to decide if a gene is differentially expressed or not... Do you know if there are some results on these topics and could you give me some references ? Thanks a lot, Olivier.