limma anova

http://www.statsci.org/smyth/pubs/limma-biocbook-reprint.pdf

http://bioconductor.org/packages/2.6/bioc/vignettes/limma/inst/doc/usersguide.pdf

https://stat.ethz.ch/pipermail/bioconductor/2006-August/013998.html

http://www.bioconductor.org/help/course-materials/2009/BioC2009/labs/limma/limma.pdf

KO <- factor(rep(1:2, each = 3, times = 2))
CELL <- factor(rep(1:2, each = 6))
design <- model.matrix(~KO + CELL)




In R:
> mat <- cbind(c(1,1,1,rep(0,9)),
+ c(0,0,0,1,1,1,rep(0,6)),
+ c(rep(0,6),1,1,1,0,0,0),
+ c(rep(0,9),1,1,1))
> colnames(mat) <- c("nu","nt","tu","tt")