[BioC] factorial experiment matrix design
Gordon K Smyth
smyth at wehi.EDU.AU
Thu May 5 14:47:25 CEST 2005
> Date: Wed, 04 May 2005 17:37:35 +0300
> From: Mikko Arvas <Mikko.Arvas at vtt.fi>
> Subject: [BioC] factorial experiment matrix design
> To: bioconductor at stat.math.ethz.ch
>
> Hi,
>
> reading the BioC list I see that things like this are asked often, but I am
> sorry I just can't get this. Could somebody show me how to do it:
>
> I have a microarray experiment with 2 strains and 3 time points from each,
> so I would like to do 2x3 factorial experiment
> to get all the genes that change in strain A (or B) during time series (10,
> 30, 60 min) and between strains in all the time points.
>
> So trying to follow limma guide, the design matrix object would be
> something like this (with 2 repeats for each):
>
> > dQex
> (Intercept) Qexa10 Qexa30 Qexa60 Qexb10 Qexb30
> 1 1 1 0 0 0 0
> 2 1 1 0 0 0 0
> 3 1 0 1 0 0 0
> 4 1 0 1 0 0 0
> 5 1 0 0 1 0 0
> 6 1 0 0 1 0 0
> 7 1 0 0 0 1 0
> 8 1 0 0 0 1 0
> 9 1 0 0 0 0 1
> 10 1 0 0 0 0 1
> 11 1 0 0 0 0 0
> 12 1 0 0 0 0 0
> attr(,"assign")
> [1] 0 1 1 1 1 1
> attr(,"contrasts")
> attr(,"contrasts")$Qex
> a10 a30 a60 b10 b30
> a10 1 0 0 0 0
> a30 0 1 0 0 0
> a60 0 0 1 0 0
> b10 0 0 0 1 0
> b30 0 0 0 0 1
> b60 0 0 0 0 0
>
> What kind of contrast (or as well design) matrix do I need to get all the
> changing genes out?
YOu need to pick out the 5 non-intercept coefficients:
cont.matrix <- rbind(0,diag(5))
Gordon
> Cheers,
> Mikko
>
>
>
> Mikko Arvas
> VTT Biotechnology
>
> Please note new tel/fax number
> from start of year 2005.
>
> e-mail: mikko.arvas at vtt.fi
> tel: +358-(0)20-722 5827
> mobile: +358-(0)44-381 0502
> fax: +358 20 722 7071
> mail: Tietotie 2, Espoo
> P.O. Box 1500
> FIN-02044 VTT, Finland
>
> Protein production group's website:
> http://www.vtt.fi/bel/bpro/proteinprod/indexe.htm
More information about the Bioconductor
mailing list