[BioC] RNA-seq analysis and two-way ANOVAs

[Gordon K Smyth]

Dear Elena,

edgeR does DE analysis at the gene or exon level, indeed any level that 
counts are summarized at.

Any package that fits genewise Poisson or negative binomial glms can claim 
to handle two-way anova, as well as more complex experiments.  The 
situation as of a little less than a year ago was reviewed in our paper on 
this topic:

  http://nar.oxfordjournals.org/content/40/10/4288

There are a few more packages now, for example QuasiSeq on CRAN.

Best wishes
Gordon

------------ original message -------------
[BioC] RNA-seq analysis and two-way ANOVAs
Elena Sorokin sorokin at wisc.edu
Thu Aug 23 23:49:50 CEST 2012

Hello,

What R packages for differential expression analysis of mRNA-seq data
are currently set up to do statistical testing with two main effects
(two-way ANOVA)?

I know that the latest versions of DESeq and edgeR can do this for
genes; I know DEXSeq can do this for exons; I don't think that the
latest versions of baySeq, DEGSeq or cuffdiff [the latter of which is
not in R, obviously] can currently do this. Would anyone in-the-know
care to update that? If there's a new software package that I missed,
please let me know about that too.

Thanks!
Elena

______________________________________________________________________
The information in this email is confidential and intend...{{dropped:4}}