[BioC] [Bioc] RNAseq less sensitive than microarrays? Is it a statistical issue?

[Steve Lianoglou]

Hi,

Some quick comments:

> - can the inability to distinguish down-regulated genes be related to
> filtering low count reads? (in order to get good separation between groups
> in an MDS plot I need to filter cpm >0.1)

It's not clear how you are filtering.

If the gene never has a cpm > 0.1 in any sample, then you'll have no
hope of detecting differential expression anywhere. If it is 0.1 in
some and 5 cpm in others, then that's more like it.

Be sure that your filtering strategy keeps the latter. You should be
filtering on the number of samples that detect the gene at a minimum
threshold and you should not be using the "label" of the samples
(which sample/treatment) for counting.

> - Is it possible that I need more coverage to improve sensitivity? I am
> currently sequencing at 50X pair end, that seemed enough. Is there any
> published study looking at RNASeq sequencing depth and sensitivity in human
> or mouse genomes?

The important number you need to look at is the number of reads you
are getting per sample.

> - Are the multiple testing corrections applied in EdgeR and DESeq  too
> stringent thus rendering the overall analysis less sensitive?

Unlikely.

> For the record my count matrices are of counts of transcripts, averaging
> counts over all exons from the same gene model for all RefSeq genes. I did
> this because the microarray data is per transcript. In log scale I have on
> average 0.7 R2 correlation between microarray intensity and RPKM from the
> same sample.

It's not clear to me how you are doing the counting, actually, but
"averaging" anything suggests to me that you are doing it incorrectly.

Look at the vignettes from a variety of packages (edgeR, DESeq2,
easRNAseq, QuasR) for suggested counting methods.

-steve

--
Steve Lianoglou
Computational Biologist
Department of Bioinformatics and Computational Biology
Genentech