[BioC] expression array normalization and different comparisons: an only "total" normalization or a normalization per comparison?

[mjonczyk]

Dear Ana,

> Dear list:
>
> I have a question about how to do expression microarray normalization 
> when
> doing group
> comparisons. I have 40 individuals (20 cases, 20 controls) and each 
> one has
> 3
> mRNA chips at different time-points. Should I normalize all the chips
> (40x3)
> together (and the subset for each comparison) or if I am comparing
> cases_time1 vs controls_time1, or cases_time1 vs
> cases_time2, should I normalize only these 40 chips involved in this
> particular comparison?  (I am working with Illumina chips)
>

If that is two colour chips you should normalize your data for each
chip separately - so called within array normalization (to correct dye 
and array specific effects), and then
normalize chips together - between array normalization (to make 
distribution of data from each chip comparable to each other, so you can 
perform statistical tests).
If this arrays are one-colour then only second normalization will be 
useful.

> I am new in the expression analysis and I would appreciate any 
> comments
> about
> advantages or disadvantages regardind these two approaches.

I **strongly** recommend you to search Illumina's pages for documents 
about their approach and reccomendations,
and then to find appropriate Bioconductor package (which implements 
recommended procedures and surely will give you sketch of workflow).
Of course literature search will be also helpful.

HTH,
Best Wishes,

Maciej Jończyk

-- 
Maciej Jonczyk,
Department of Plant Molecular Ecophysiology
Faculty of Biology, University of Warsaw
02-096 Warsaw, Miecznikowa 1
Poland



-- 
This email was Anti Virus checked by Astaro Security Gateway. http://www.astaro.com