Hi Michael,

I am having hard time with one of the replicates in a three replicate experiment. This is a control sample. But the other replicates look good on PCA plot and also on pairwise scatter plot comparison. Can I ignore this one replicate and perform the deseq2 analysis. I understand that the tool will allow and it will loose the statistical power, but my question is does this have a significant effect on the DE genes?

Our experimental collaborators seem fine with it because this kind of variation is anticipated but I am not convinced to do DE analysis with variation in one out of three samples. 

Whats your experience on this?

Greatly appreciate your feedback.

-Regards

Prasad

Let's continue the discussion on the original thread to keep things organized. This is preferable to starting many new threads, if the topic is the same.

deseq2 - PCA plot