[BioC] A doubt with contrasts in limma

[James W. MacDonald]

Hi Davide,

On 8/11/2012 7:51 AM, Cittaro Davide wrote:
> Hi all,
> If I have a 2x2 design (gen[AB], treat[CD]), I've created a design matrix so that I have 4 classes: A.C, A.D, B.C and B.D. Now I'm interested in finding the global differences between genA and genB, the global differences between treatC and treatD and their interaction, is it correct to specify contrasts like:
>
> (A.C + A.D) - (B.C + B.D) for genotypes
> (A.C + B.C) - (A.D + B.D) for treatments

The above two are not really what you want. Instead you should divide 
each group by two, e.g.:

(A.C + A.D)/2 - (B.C + B.D)/2 for genotypes

Best,

Jim



> (A.C - A.D) - (B.C - B-D) for interactions (as in the manual)
>
> ?
>
> d
> --------------------------------------------------------------------------
> LA TUA CURA E' SCRITTA NEL TUO DNA. AL SAN RAFFAELE LA STIAMO REALIZZANDO.
> AIUTA LA RICERCA, DAI IL TUO 5XMILLE - CF: 03 06 80 153
> info:www.5xmille at hsr.it - www.5xmille.org
>
> Disclaimer added by CodeTwo Exchange Rules 2007
> http://www.codetwo.com
>
> _______________________________________________
> Bioconductor mailing list
> Bioconductor at r-project.org
> https://stat.ethz.ch/mailman/listinfo/bioconductor
> Search the archives: http://news.gmane.org/gmane.science.biology.informatics.conductor

-- 
James W. MacDonald, M.S.
Biostatistician
University of Washington
Environmental and Occupational Health Sciences
4225 Roosevelt Way NE, # 100
Seattle WA 98105-6099