[BioC] edgeR glm

[Steve Lianoglou]

Hi,

On Tue, Apr 5, 2011 at 5:04 PM, Shreyartha Mukherjee
<shreyartha at gmail.com> wrote:
> Hi Gondon,Bioconductor experts,
>
> I have RNA-seq data for 3 genotypes(each genotype having 3 biological
> replicates) and 30,000 genes.
> I was trying to find out differentially expressed genes across all
> genotypes. Here is the code I am using.
>
>  library(edgeR)
> y<-read.table('test_12339.txt'
> )
>  d <- DGEList(y, group = rep(1:3, each = 3), lib.size = lib.sizes)
>  d <- calcNormFactors(d)
>  times <- rep(c("p1","p2","f1"),c(3,3,3))
>  times <- factor(times,levels=c("p1","p2","f1"))
>  design <- model.matrix(~factor(times))
> disp <- d2$tagwise.dispersion
>  fit <- glmFit(d,design,dispersion=disp)
>  lrt <- glmLRT(d,fit)
> topTags(lrt)
>
> Does this code provide me with tags differentially expressed across all
> conditions? (I am not looking for pairwise differential expression
> but across all conditions)

Out of curiosity, what does it mean for a gene to be differentially
expressed in all (three) conditions?

I mean -- what would its expression pattern look like across your
three genotypes?

If G1, G2, and G3 are your three different genotypes, are you looking
for a gene A that's differentially expressed when you compare G1 vs.
G2 AND G2 vs G3 AND G1 vs G3 (or something(?))

-- 
Steve Lianoglou
Graduate Student: Computational Systems Biology
 | Memorial Sloan-Kettering Cancer Center
 | Weill Medical College of Cornell University
Contact Info: http://cbio.mskcc.org/~lianos/contact