[BioC] Factorial design with two-channel microarray data with Limma

[Gordon K Smyth]

Dear Eran,

You have to read the limma User's Guide.  Section 8.1.1 tells you what to 
do exactly.  Your experiment is so simple you don't even need design or 
contrast matrices.

Best wishes
Gordon

> Date: Thu, 01 Oct 2009 11:38:17 -0400 (EDT)
> From: "eran tauber" <etauberle1_7rh at lycos.com>
> Subject: [BioC] Factorial design with two-channel microarray data with
> 	Limma
> To: <bioconductor at stat.math.ethz.ch>
> Content-Type: text/plain
>
> Hello Bioconductor users,
>
> We have carried the following experiment: two different fly strains (58 
> and 104) were tested (each) under two different conditions, Control and 
> Pulse (C and P). For each strain, samples from two conditions were 
> labelled with Cy3 and Cy5 and hybridized together to the same array. Two 
> (biological) replicates were carried for each hybridization.
>
> We are interested in finding those genes for which the difference C-P is 
> maximized (and significant) between the strains. In other words, we are 
> interested in the difference of the differences: (C[58]-P[58]) – 
> (C[104]-P[104]).
>
> The target file looks like that:
>
> SlideNumber            FileName       Cy3     Cy5
>
> 1          58 UP1_LP1.gpr       control            pulse
> 2          58 UP2_LP2.gpr       control            pulse
> 3          104 UP1_LP1.gpr     control            pulse
> 4          104 UP1_LP1.gpr     control            pulse
>
> Could anyone please help us with writing the design and contrast matrix 
> for analysing this experiment with Limma?
>
> Many thanks, Eran
>
> Eran Tauber (PhD)
> Lecturer in Molecular Evolution
> Dept. of Genetics
> University of Leicester
> University Rd, Leicester LE1 7RH
> England
> ____________________________________________________________
> Phone: 44 (0)116 252-3455, 252-3421 (lab) Fax: 44 (0)116 252-3378
> www.le.ac.uk/genetics/et22