[BioC] edgeR - one way ANOVA, coef parameter interpretation in glmLRT()

[Michal Okoniewski]

Hey Mark,

A question on one-way ANOVA design with edgeR:

I do something like that for 3 groups, so similar case like in the vignette (glm functionality):


d <- calcNormFactors(d)

d <- estimateCommonDisp(d, verbose=TRUE)

d <- estimateTagwiseDisp(d)


TS <- factor(cc)

design<- model.matrix(~0+TS)

colnames(design) <- levels(TS)


 fit <- glmFit(d, design)

 fit2 <-  glmLRT(fit)

 # or      fit2 <-  glmLRT(fit, coef=1) ???


 outTab <- topTags(fit, n=30666)

An the question is:
__ How to get the p-values of the F-test for one-way anova? In a basic comparison of 3 groups… __

In the vignette you have written:

"The fi t has three parameters. The fi rst is the baseline level of group 1. The second and third
are the 2 vs 1 and 3 vs 1 di ferences." edgeR vignette (p17, date March31st 2013)

This sounds to me a bit like hard-coded contrasts… in addition, default value of coef is ncol(glmfit$design), so 3 in this case. Shall I use coef=1?
I am a bit confused…

Thanks!

I send a copy to BioC list, as I could not find the answer there.
If I ask stupid question – please give me a simple solution too :)

Cheers,
Michal


> sessionInfo()
R version 3.0.1 (2013-05-16)
Platform: x86_64-apple-darwin10.8.0 (64-bit)

locale:
[1] en_US.UTF-8/en_US.UTF-8/en_US.UTF-8/C/en_US.UTF-8/en_US.UTF-8

attached base packages:
[1] stats     graphics  grDevices utils     datasets  methods   base

other attached packages:
[1] edgeR_3.2.3  limma_3.16.5

loaded via a namespace (and not attached):
[1] tools_3.0.1



--
Dr. Michal Okoniewski

Functional Genomics Center Zurich (mostly Wed-Fri)
Winterthurerstrasse 190, 8057 Zurich
Room Y32 H66, Phone: +41 44 635 39 24

Department of Neuroimmunology and MS Research,
UniSpital Zurich, (mostly Mon-Thu)
Raemistrasse 100, 8091 Zurich
Room: OPS D37, Phone: +41 44 255 1756