Group sizes in RNA-seq DE

[moriah]

Hi all,

I would like to run an RNA differential expression analysis on two groups (two cancer subtypes) but my number of samples in the groups is not even close to equal.
I have one group with about 120 samples and the other one has around 25 samples, is it O.K to run DEseq or EdgeR on such different size groups ? Would the possible bias be corrected somehow ? If not, What is the maximal group difference I can work with in DE?

Thanks,

Best,
Moriah

[Will Nelson]

any answer?

Moriah,

did you ever get an answer? I'm seeing that EdgeR actually generates errors and doesn't work for unequal samples. Is this a fundamental limitation?

Will Nelson

[Simon Anders]

Having different group sizes in no problem, neither for DESeq nor for edgeR.