[BioC] affy scalevalue for RMA?

[Celine Carret]

Dear list,

I'm trying to better understand  what the "affy.scalevalue.exprSet" 
function does exactly (affy package). I know it is normally used for MAS 
pre-processing, but I found it very useful in an RMA context, to kind of 
normalise different set of chips to the same "subset" that has been 
median polished 1st then used as a scale factor.
When looking at the function in a prompt shell, I read "mean" although I 
thought this function was using the scale factor to scale up the medians 
and not the means. I'm not sure it would make a big difference anyway, 
but I just wondered, if I medpolish my subset, wouldn't it make more 
sense to continue using the medians? I noticed on the boxplots before 
and after the scaling that the intensities distribution vary and tend to 
be more equals on the upper limit of the quartiles after scaling up the 
chips, which is not what RMA does, or does it?

Thank you very much for your time
Celine


-- 
Celine Carret PhD
Pathogen Microarrays group
The Wellcome Trust Sanger Institute
Hinxton, Cambridge CB10 1SA, UK.
tel. +44 (0)1223 494 940
fax. +44 (0)1223 494 919
email: ckc at sanger.ac.uk
http://www.sanger.ac.uk/PostGenomics/PathogenArrays/