Hello, I would like to create a design matrix for an experiment but I am a bit confused and I would appreciate to have some expert views. We are talking about dye-swaped microarrays (chicken genome, but it doesn't matter). We have two conditions : Vasa can be + or - and SF absent or present first set of arrays(1 and 2): Vasa+/vasa- and the dye-swap (no SF at all) second(3 and 4): vasa-/vasa- SF and the dye-swap third(5 and 6): vasa-/vasa+ SF and the dye-swap fourth(7 and 8): vasa- SF/vasa+ SF and the dye-swap can I integrate this into one model matrix ? I have done the analysis separating the arrays because the only comparaisons that interested the biologists were 1/2, 3/4, 5/6 and 7/8 but I don't think this is very nice and would like to make it better. Thank you Aur?lie -- Aur?lie Laugraud PRABI B?t.Mendel 2? ?tage 16 rue Raphael Dubois 69622 Villeurbanne Cedex tel: 0472432826