Important! Template update for nf-core/tools v3.4.1

[nf-core-bot]

Version 3.4.1 of nf-core/tools has just been released with updates to the nf-core template. This automated pull-request attempts to apply the relevant updates to this pipeline.

Please make sure to merge this pull-request as soon as possible, resolving any merge conflicts in the nf-core-template-merge-3.4.1 branch (or your own fork, if you prefer). Once complete, make a new minor release of your pipeline.

For instructions on how to merge this PR, please see https://nf-co.re/docs/contributing/sync/.

For more information about this release of nf-core/tools, please see the v3.4.1 release page.

[MaximilianStammnitz]

@BenjaminWehnert1008 would be great if you can start looking at these latest nf-core template updates and alter the pipeline files accordingly (I'll mostly work on completing and ironing our documentations today and tomorrow)

[BenjaminWehnert1008]

Hi @mashehu & @MaximilianStammnitz,

This PR appears to be based on an outdated version of our repository, not the current dev branch.
For example, renaming a file from dmscore.nf to deepmutscan.nf. This renaming was already completed and merged into dev a long time ago, when we changed the pipeline name from dmscore to deepmutscan. Most other changes (maybe even all) also refer to outdated files.
What is the best way to resolve this? Can we close this PR and try to trigger a new, fresh template sync?

Thanks!

[BenjaminWehnert1008]

I think it's based on master branch instead of dev branch...

[MaximilianStammnitz]

@BenjaminWehnert1008 @mashehu we have been working from the dev branch all these months. Looks like it's the right time now to merge all of our updates into master? This should facilitate any downstream work on the nf-core template liftover, remaining conflicts with linting, etc.

I've just created the PR to do this, but would be good @mashehu if you can confirm that's the right way (in line with: https://nf-co.re/docs/tutorials/adding_a_pipeline/first_release): #17

[mashehu]

Looks like it's the right time now to merge all of our updates into master?
Nope, we only have released versions on master/main. so dev->master PRs will always be release PRs with special checks (and two review requirements. I will try to look later today into why the sync seems to have picked the wrong branch here.

[MaximilianStammnitz]

Great, thanks @mashehu. Perhaps the branch choice is a shared underlying issue between both the documentation update and what BenjaminWehnert1008 has seen.

[mashehu]

the sync action used the dev branch https://github.com/nf-core/tools/actions/runs/19279454212/job/55127082953 🤔

[mashehu]

I think this sync actually worked fine, I will resolve the merge conflicts locally so you can see if the changes are fine

[mashehu]

for example, renaming a file from dmscore.nf to deepmutscan.nf.

I didn't see that in this update. where was it?

quick points I saw while just looking at merge confilcts:

• don't define default parameters in deepmutscan.nf, we use nextflow.config for that
• why this R functions as channel setup? See for example differentialabundance for a pipeline with a lot of r modules
• I would recommend using (local) subworkflows to structure your deepmutscan.nf a bit more
• you should move your test data over to https://github.com/nf-core/test-datasets/

[BenjaminWehnert1008]

Hey @mashehu thank you very much for this feedback, I wasn't aware of the opportunity of using a templates folder and command for the R scripts. This makes much more sense than channeling! Also, the other points all sound valid. Gonna take a short while for me to change all of this, gonna come back to you asap:)

[BenjaminWehnert1008]

Hey @mashehu & @MaximilianStammnitz, finally coming back to you:) I opened two PRs to establish all changes requested. Now there's no channeling of R scripts anymore (instead using templates). I also implemented a subworkflow for the fitness calculation, which cleans up the main script. What are your thoughts on that?

I tried to merge the test data (https://github.com/BenjaminWehnert1008/test-datasets/tree/deepmutscan) into the nf-core repo, but I wasn't sure how to add a new branch for our pipeline. Do I have the rights to do it? (still bit new to all this stuff...)

Thanks:)

[mashehu]

I added you to the nf-core github org (usually done via the #github-invitations slack channel), and you should now have all the right permissions to set up a branch on the test-dataset repo: here is a step-by-step tutorial on how do to it: https://github.com/nf-core/test-datasets/blob/master/docs/ADD_NEW_DATA.md

I had a look at the other PRs, good progress!

[MaximilianStammnitz]

@BenjaminWehnert1008: for the test data, can we try to use the latest set of 2 x input + 2 x output mini-fastqs (50k reads each)? Files are currently in our repo folder: https://github.com/nf-core/deepmutscan/tree/dev/testdata

[MaximilianStammnitz]

@mashehu our Deep Mutational Scanning test dataset has an associated fasta file reference – but here, this is not a species genome, only a small test gene from Arabidopsis (~1.3 kb). Is that okay or should I ask in the Slack?

[MaximilianStammnitz]

Also @mashehu, is it worth merging this branch now? Just to also run some local tests on this latest version with the v3.4.1 nf-core template and @BenjaminWehnert1008's latest changes

[mashehu]

@mashehu our Deep Mutational Scanning test dataset has an associated fasta file reference – but here, this is not a species genome, only a small test gene from Arabidopsis (~1.3 kb). Is that okay or should I ask in the Slack?

looks small enough. 🙂

[mashehu]

Also @mashehu, is it worth merging this branch now? Just to also run some local tests on this latest version with the v3.4.1 nf-core template and @BenjaminWehnert1008's latest changes

sure, while you are still developing it is fine if not all the CI is passing