[BioC] design matrix, again

[Giovanni Coppola]

Hi all,
as many before, I have hard time in modeling a design matrix for a 24-slide cDNA experiment, including 4 experimental animals (each with 3 regions studied), 4 controls, and a dye swap! Thus, no common reference and no before/after data (I saw these kinds of matrices previously posted), and many possible levels of analysis (by animal, by region, or by experimental condition). Did anyone deal with something like this?
In addtion, can I ask to Gordon Smyth to make public the reply to Pete (http://files.protsuggest.org/biocond/html/3159.html)? I am sure this will be a good starting point, to me and others. 
Thank you in advance
Giovanni

	[[alternative HTML version deleted]]