[DylanMHouse]

Adds a new module for longphase/modcall. This tool is commonly used for calling allele-specific modifications in long-read Nanopore and PacBio reads, given the aligned BAM file preserved the MM,ML tags with these modifications. The user can also optionally input a variant call file containing SNV information.

Inputs for this module is a meta-map in the format [meta, bam, bai, snv], and the reference [meta2,fasta,fai]

Output is a compressed modcalled generated vcf which can then be provided as an input to longphase/phase for phasing.

Tests cover if all inputs are provided, and another if the snv file is omitted.

PR checklist

• [x ] 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.
• [ x] Remove all TODO statements.
• [ x] Broadcast software version numbers to topic: versions - See version_topics
• [ x] Follow the naming conventions.
• [ x] Follow the parameters requirements.
• [ x] Follow the input/output options guidelines.
• [ x] Add a resource label
• [ x] 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:
    • [x ] nf-core modules test <MODULE> --profile docker
    • nf-core modules test <MODULE> --profile singularity
    • [ x] nf-core modules test <MODULE> --profile conda

[github-actions]

This PR exceeds the recommended size of 1000 lines. Please make sure you are NOT addressing multiple issues with one PR. Note this PR might be rejected due to its size.

[vagkaratzas]

Thanks for the module, the structure and test data usage look good. Blocking points:

• The container directive was copied from longphase/phase (longphase 2.0.1, plus samtools) and does not match environment.yml (longphase 2.1.1, htslib) or the containers: block in meta.yml; the snapshots were generated with the wrong version. Please run nf-core modules containers create longphase/modcall, then regenerate the snapshots.
• htslib is in environment.yml and bgzip is called, but there is no versions_htslib emit; the longphase version is also the full usage text rather than a bare 2.1.1.
• meta.yml still has the template description, no args_id, and incomplete ontologies and keywords. Test names should follow the <organism> - <inputs> - <outputs> pattern.

While you are here (not blocking):

• modules/nf-core/longphase/modcall/main.nf:49-52: trailing blank lines can be removed.

Not run: G1 and G2. The PR's CI had not run the lint or nf-test jobs when reviewed (only the size label and mergeability checks were present), and the local runs were stopped before finishing, so please make sure nf-core modules lint longphase/modcall and the nf-test CI jobs pass. G3 (nextflow lint) passed.

Diff 8.21/10 · Overall 8.21/10
nf-core-reviewer 0.3.0 / vagkaratzas-review (rubric v2) · claude-sonnet-5-5

[vagkaratzas]

The container directive does not belong to this module's environment: it is the longphase_samtools image copied from longphase/phase (the snapshot reports longphase 2.0.1 while environment.yml pins 2.1.1), it does not match the containers: block in meta.yml, and the conda lock file contains no htslib. Please run nf-core modules containers create longphase/modcall after finalising environment.yml, so the directive, meta.yml and lock files all describe the same build, then regenerate the snapshots.

[vagkaratzas]

Placeholder description left over from the template. Please describe the module in one short line, e.g. Call allele-specific base modifications from a long-read BAM with MM/ML tags.

[vagkaratzas]

CI only runs automatically for nf-core members. Please ask for a GitHub invitation in the nf-core Slack #github-invitations channel; once you've joined, CI will run on your next push.

Comment added by Claude

[DylanMHouse]

Thank you for the comments. I have updated the tests according to the template, and I now use the correct container for the updated versions. The tests are also regenerated. I found an EDAM for .bai format but no EDAM format was found for .fai so I left that blank.

I will ask for the GitHub invitation on Slack.

Thanks again, I am still new to this!

[vagkaratzas]

• Blocking: under the default prefix the output can take the name of the SNV input (${meta.id}.vcf.gz, as clair3 writes it), which makes bgzip fail or truncates the upstream file. A _modcall suffix on the default prefix fixes it.
• Small: the longphase version in the snapshot is LongPhase Ver 2.1.1 instead of 2.1.1 (my earlier sed was wrong), and the BAI ontology needs its # BAI comment.

Thanks for working through the last round. The container, the htslib version, meta.yml and the test structure all look good now, and CI passes on all three profiles.

While you are here (not blocking):

• modules/nf-core/longphase/modcall/tests/main.nf.test:37-38,68-72: indentation is off and there are extra blank lines; main.nf:9-10 has a double blank line.

Diff 9.11/10 · Overall 9.11/10
nf-core-reviewer 0.5.0 / vagkaratzas-review (rubric v4) · claude-opus-5-5

Comment added by Claude Opus 5.5

[DylanMHouse]

Alright, I have added the changes you suggested and I think the indentation is fixed, and the snapshots were regenerated. Strangely enough, I had to link the module on my local computer because the prettier hook failed to install on the cluster. xD.

[vagkaratzas]

• Blocking: the stub default prefix does not match the script's (_modcall suffix missing), so stub snapshots carry the wrong file name. After fixing, regenerate the two stub snapshots.
• The previous blockers are resolved: the prefix suffix avoids clobbering the SNV input, versions are bare (2.1.1, 1.24), the BAI ontology comment is there, and CI passes (lint and nf-test on the current head). Every modcall file option (-b, -r, --snp-file, -o) is wired.

While you are here (not blocking):

• modules/nf-core/longphase/modcall/main.nf:9-10: double blank line, still there from last round.
• modules/nf-core/longphase/modcall/tests/main.nf.test:28,38-39,67-72: indentation of the closing ]/} is off and there are extra blank lines, still there from last round.
• modules/nf-core/longphase/modcall/meta.yml: $args2 goes to bgzip but htslib has no entry under tools:; add one with args_id: "$args2".

Diff 9.64/10 · Overall 9.64/10
nf-core-reviewer 0.5.0 / vagkaratzas-review (rubric v4) · claude-sonnet-5-5

Comment added by Claude Sonnet 5.5

[vagkaratzas]

• The stub default prefix now matches the script (${meta.id}_modcall), and the stub snapshots carry the right file name. Lint, nf-test and nextflow lint all pass on the current head.

Approval should be on its way soon.

While you are here (not blocking):

• modules/nf-core/longphase/modcall/meta.yml: $args2 goes to bgzip, but htslib has no entry under tools:; add one with args_id: "$args2".
• modules/nf-core/longphase/modcall/main.nf:8-9: double blank line.
• modules/nf-core/longphase/modcall/tests/main.nf.test:33-35,59-63,23: closing ]/} indentation is off and there are extra blank lines.

Diff 10.00/10 · Overall 10.00/10
nf-core-reviewer 0.5.0 / vagkaratzas-review (rubric v4) · claude-sonnet-5-5

Comment added by Claude Code

[awgymer]

Ok, you can attempt to update this PR with main and then rerun the CI. However I would personally like to see the longphase containers/versions unified across the modules so that we don't have multiple different versions of a tool across the different subtool modules

[vagkaratzas]

and run the latest version of nextflow lint to pass pre-commit :D

[DylanMHouse]

I am using the same container/version used in longphase/phase so I wlll keep that updated.