[BioC] tech reps revisited in limma

Simon Melov smelov at buckinstitute.org
Sun May 2 22:40:28 CEST 2004

After viewing the various postings on this topic ("technical replicates 
(again!): a summary"), I couldnt find a solution to a pretty basic 
problem. I have two technical replicates per sample, and 7 samples per 
treatment. I have 4 treatments, which I want to contrast with each 
other (all using 2-color chemistry).

I created a design which treated each sample as a duplicate (i..e 1, -1 
as dye swaps were involved), and then tried to use this to create 
contrasts, with subsetting etc. This ended up being very laborious, and 
there may not be a way to do this easily in limma, but I could be 
missing something.

What I have is this: 4 treatments: A,B,C,D
For each treatment, I have 2 technical replicates  (dye swaps), I want 
to average these in limma, and then look at comparisons between groups, 
and create contrasts between A and B, etc. However, I dont see an easy 
way of doing this, short of a lot of subsetting, and renaming 

Any help would be appreciated.



More information about the Bioconductor mailing list