Hi all, just a quick question: Having read Ben's comment on normalization (only few genes should change), I came to think of Affymetrix concept of using 100 genes which don't change much as a normalization strategy. Has anybody looked into this / tried some R implementation? Cheers, hinrich d8-)