[BioC] Limma design matrix

[Gordon Smyth]

At 06:04 AM 27/07/2004, Yihuan Xu wrote:
>Hi,
>
>I am using Limma for a non-affy single channel dataset. the design is a 2X2
>factorial design, but one chip is missing.
>
>array   gender   genotype
>1        M            mutant
>2        M            mutant
>3        F             mutant
>4        F            mutant
>5        M            wild
>6        M            wild
>7        F             wild
>
>How to make the design matrix?

Why not

design <- model.matrix(~factor(gender)*factor(genotype),data=targets)

where 'targets' is your targets data.frame?

Gordon

>Thank you very much.
>
>Yihuan