[Bioc-sig-seq] RNAseq stat analysis

[Eric Bonnet]

Dear R-gurus,

We have mRNA expression data that was generated by solexa-illumina deep 
sequencing (75 nt reads), for two developmental stages, and we would 
like to have genes that are differentially expressed, by using 
procedures that were developed for microarrays, like samr, siggenes or 
limma.

Counts for unique reads are normalized by gene length and also by the 
total number of reads obtained for the library, followed by a log2 
transformation.

My problem is that many genes have a count of 0, so taking the log is 
not possible. The workaround I used is to add arbitrarily 1 to all the 
counts before the log transformation.

1/ Does that sound ok to you guys?

2/ Does anyone knows a package that would deal directly with the counts 
for the statistical analysis?


Thanks a lot,

Best regards,

-- Eric