[BioC] DESeq2 high fold change values for comparison of samples with zero counts

[Simon Anders]

Hi Noa

On 08/01/14 14:11, Noa Henig wrote:
> I would expect, for example:
> genes having these counts T0: 29852 and 39519; T1: 37352 and 44385;
> T3:52539 and 58,556.
> The log2 fold change of T0/T1 was 3.28 (p-adjusted was 0.04), while the
> log2 fold change of T2/T0 was 2.96 (p-adjusted was 0.105)

Please tell us the values of the size factors (run "sizeFactors(dds)")
to see how far off this is.

And please always post your code, not just the output.

  Simon