I would like to use the best procedure to merge Affy U133A and U133B ( with RMA normalization) from different sets on the same samples. Have you some ideas ? Philippe