Hello, I am analyzing the GNF/Novartis data (Luscombe re-analysis E-TABM-145). This experiment contains samples from a number of organism parts. My aim is: 1. to compare the gene exprssions in a single organism part to the average expression in all samples (the grand mean). 2. to compare the gene expessions in each brain region to the average expression in all brain samples. For both these analyses I need design matrices and I frankly have no idea how to construct them (I couldn't find them in tutorials or mailing lists). I would greatly appreciate advice, since I can't afford mistakes with these matrices. Thanks in advance, Hans Koolschijn Groningen The Netherlands [[alternative HTML version deleted]]