[BioC] limma stdev.unscaled

[Jason Skelton]

Hi

I wanted to calculate coefficients of variation for some genes using 
data analysed with limma
(two-colour  microarray data), and  found that the SD for all genes in 
each comparison were the same
e.g.  from fit$stdev.unscaled

Comparsion 1       Comparison 2    Comparison 3
 Min.   :0.5599   Min.   :0.5599   Min.   :0.7919
 1st Qu.:0.5599   1st Qu.:0.5599   1st Qu.:0.7919
 Median :0.5599   Median :0.5599   Median :0.7919
 Mean   :0.5599   Mean   :0.5599   Mean   :0.7919
 3rd Qu.:0.5599   3rd Qu.:0.5599   3rd Qu.:0.7919
 Max.   :0.5599   Max.   :0.5599   Max.   :0.7919

I understand that eBayes uses a pooled variance approach but these 
values are obtained before eBayes was used.
Could anyone help explain the values I am getting please ?

Cheers

Jason

-- 
--------------------------------
Jason Skelton
Pathogen Microarrays
Wellcome Trust Sanger Institute
Hinxton
Cambridge
CB10 1SA

Tel +44(0)1223 834244 Ext 7123
Fax +44(0)1223 494919