Hello,
For my CRISPR screen, I have three conditions: control, and treatment A and B. I analyzed these samples in one pinapl run. And it gives me the wrong average gene and sgRNA rankings for B (the results for the single replicates are fine). Namely, it seems to include samples from A for the averaging (it then seems I have the same hits for both conditions, which is definitely wrong). The same thing already happened some months ago when I first analyzed the experiment, and now happened again when I performed a re-analysis.
I'm 100% sure, I annotated the samples correctly.
There seems to be something wrong.
Best wishes,
Maria