Hello everyone! I have a question about BAGEL2 analysis in crisprseq. I have two biological replicates of two samples and this is my contrast file

reference;treatment

Day0_1,Day0_2;Untreated_1,Untreated_2

When I perform the analysis with nf-core/crisprseq (2.4.0dev), BAGEL2 calculates fold change by using reference column Day0_1 (not Day0_2). But if I do this analysis with nf-core/crisprseq (2.3.0), the fold changes are instead calculated using the mean of Day0_1 and Day0_2 as the reference. I couldn't decide which one is true (the mean-based approach seems incorrect to me) Could this be a bug? How should I fix this? Many thanks.

Hi, this could definitely be a bug... in recent commits on the dev branch I had to refactor the contrast definition to make sure that sample name appear in the mageck quantification (see https://github.com/nf-core/crisprseq/pull/252), maybe I introduced a bug in these changes I need to look into this, give me a couple of days. It would be super-helpful if you could open an issue, so that I do not lose track of this issue.

No worries, Matteo. I am actually asking this because I am confused with handling the biological replicates. It shouldn't be a mean of the replicates as a reference column right? I will also open this as an issue as you requested.