Hello there,

Online I can find most examples related to cancer datasets with a lots of samples. However, I have a setup of only 108 samples on FPKMs, distributed into 2 conditions (cond1), 6 further conditions (cond2) each, 3 time-points and each 3 replicates. Can DESeq2 handle this low sample number and replicates as well, and give me DEGs for cond1 and time, while cond2 should be constrained to not be differentially expressed? Is there a lowest number of samples limit?

Thank you

There is not a low number of samples limit as long as you have positive degrees of freedom.

Note that you cannot use FPKM but DESeq2 requires counts.

while cond2 should be constrained to not be differentially expressed

See lfcThreshold as described in the vignette or the 2014 paper.