Repository navigation
hanging when finding genewise dispersions #453
Description
Hello pydeseq2 devs.
Thank you all for developing this easy-to-use analysis tool. I'm running into extremely long runtimes (or possibly just hanging) when running dds.fit_genewise_dispersions(). I've seen some chatter about similar issues in the past when running dds.deseq2().
As others have pointed out, this issue appears linked to occurences of 0 within the matrix. Adding 1 to all counts seems to resolve the issue, but I'd prefer to avoid this if possible.
Does anyone have practical advice for working around this issue?
Activity
Sizerta commentedon Oct 17, 2026
I tracked down the bottleneck: most of the time is spent fitting size factors iteratively, which
fit_genewise_dispersions()triggers when every gene has at least one zero. I’m working on a PR to make this faster.