@Nico Trummer the PR for Cell Cycling is ready! https://github.com/nf-core/scdownstream/pull/266

Should I update the snapshots in

test

for you, as I have access to HPC?

@Nico Trummer ah, i just thought that there is no need to update them, as cell cycling don't emit additional files in the final workflow output

Okay I did not mention before, but the version captures also play a role

Ah, okay, I'll do it now, if any problems I will ask help!

Technically not all of them, you can check the CI for the ones that actually fail But I guess you can run all of them as well I created the

refresh.sh

script for updating all of them in parallel, you can use it if it helps you

@Nico Trummer I have a question about the profile, we should use for testing, described in the comments to PR https://github.com/nf-core/scdownstream/pull/266#issuecomment-4060286794

Hey, saw your question, but I'm currently traveling and can't look into it If the singleR test relies on some param that is part of the test profile, then this would make sense, but it would not be optimal as the tests should be 'standalone'

As far as I can see from github logs, the testing there is carried out with testing profile (but maybe this doesn't mean it runs with --profile test). But if github run it without test profile, how does it get the result, that we cannot reproduce on different local systems👀

In general this should be handled by the .nf-test.config (or similar) in the repo root

@Nico Trummer yeah, as far as I can see, this is exactly the main reason (absent of '+'), I have a different snapshot. celldex_reference is not defined within .nf.test files for pipeline testing. On the other hand, nf-test.config (with is used while nf-test test) indicated test profile in it, so celldex_reference is being used from conf/test.config.