Anarcii

[Clara0611]

nf-core/modules pull request

PR checklist

Closes #10805

Adds a new module: ANARCII for auto-numbering TCRs and antibodies using a language model
Tool github: https://github.com/oxpig/ANARCII

• This comment contains a description of changes (with reason).
• If you've fixed a bug or added code that should be tested, add tests!
• If you've added a new tool - have you followed the module conventions in the contribution docs
• If necessary, include test data in your PR.
• Remove all TODO statements.
• Broadcast software version numbers to topic: versions - See version_topics
• Follow the naming conventions.
• Follow the parameters requirements.
• Follow the input/output options guidelines.
• Add a resource label
• Use BioConda and BioContainers if possible to fulfil software requirements.
• Ensure that the test works with either Docker / Singularity. Conda CI tests can be quite flaky:
  • For modules:
    • nf-core modules test <MODULE> --profile docker
    • nf-core modules test <MODULE> --profile singularity
    • nf-core modules test <MODULE> --profile conda

[Clara0611]

Tests were changed to only assert versions and the run-independent portion of the header line, as the computed scores differ slightly between architectures (observed here from the fourth decimal onward). This was noted by the developers of anarcii as well: "[...] However, we have observed that they show minor variation on different architectures and versions of python/torch in a small number of sequences. These differences are minimal." (source: https://github.com/oxpig/ANARCII/wiki/FAQs#understanding-sequence-scores, retrieved 30.03.26).

[mashehu]

maybe you can make this assertion easier to read using nft-csv and checking the header? https://github.com/lukfor/nft-csv#columnnames

[mashehu]

please don't use an https instead of an oras url here: nf-co.re/docs/tutorials/nf-core_components/using_seqera_containers

[Clara0611]

Thanks for reviewing, both done :)

[famosab]

Did you already manage to add the test-data to the test-datasets repository? :)

[mashehu]

Suggested change

	def anarciiResults = path(process.out.anarcii.get(0).get(1)).csv
	assert "Name" in anarciiResults.columnNames
	assert "Chain" in anarciiResults.columnNames
	assert "Score" in anarciiResults.columnNames
	assert "Query start" in anarciiResults.columnNames
	assert "Query end" in anarciiResults.columnNames
	assert path(process.out.anarcii[0][1]).csv.columnNames == ["Name", "Chain", "Score", "Query start", "Query end"]

or are the optionally more column names?

[Clara0611]

There are 128+ other columns corresponding to the numbering, which depends on the dataset. I can find out which exact numbering (includes alternatives sometimes, like 112A, 112B etc) this test data set produces, but I thought that including all of those would be hard to read and not representative of the tool, only of this specific test dataset. However, if you feel I should include them, it is possible.

[famosab]

Are you still struggling with instable md5sums even after updating the test data?

[Clara0611]

Yes, unfortunately while the scores look better, they are still not 100% stable across systems

[famosab]

Are you still struggling with instable md5sums even after updating the test data?