Add vepyr/annotate module

[mwiewior]

Add vepyr/annotate to annotate VCFs against a vepyr Parquet cache and emit a bgzip VCF with its tabix index. The module uses Bioconda vepyr 0.7.0 and htslib 1.24, accepts reference indexes and an optional plugin cache, and uses task CPUs for annotation parallelism with a single-worker fallback for unindexed input.

The real annotation test uses 1,000 HG002 chr22 records, the Ensembl 116 cache and GRCh38 reference now published by nf-core/test-datasets#2270. It extracts the cache through UNTAR and snapshots the --everything annotation digest, output names and tool versions. A second test covers stub output. Docker/Singularity container metadata and generated Conda lock files are included for both supported architectures.

Validation in a fresh checkout of nf-core/modules:

• nf-core modules lint vepyr/annotate: 80 passed, 0 warnings, 0 failures.
• Repository pre-commit hooks: passed, including schemas, formatting and Nextflow lint.
• nf-test test modules/nf-core/vepyr/annotate/tests --profile=+docker,arm64 --ci: 2/2 passed on Nextflow 25.10.2 (the CI version) and 26.04.6, with nf-test 0.9.5 and no test-data URL override.
• The published-data output also matches the recorded Ensembl VEP 116 record-body MD5 1d4a92b815eb6193e6f4546f4ab35978 for all 1,000 records.

Snapshots were generated from the published data on ARM64 and remained unchanged during both upstream test runs. Native x64, Conda and Singularity execution still need upstream CI validation. This PR contains only the module; the normalisation subworkflow will follow separately after the module is merged.

Generated by Codex

PR checklist

• 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. Published in Add test data for the vepyr/annotate module (Ensembl 116, HG002 chr22) test-datasets#2270.
• Remove all TODO statements.
• Broadcast software version numbers to topic: versions.
• 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:
  • Docker: equivalent direct nf-test command above, using the ARM64 profile.
  • nf-core modules test vepyr/annotate --profile singularity
  • nf-core modules test vepyr/annotate --profile conda

[SPPearce]

Can you please join the nf-core organisation via the github-invitations channel on slack

[SPPearce]

This needs to be explicitly incorporated into the input channel then.

[mwiewior]

Done in 69cbfa9: the reference tuple is now tuple val(meta3), path(fasta), path(fai), path(gzi), with gzi documented as its own optional input in meta.yml ([] for an uncompressed FASTA). Tests updated; outputs and the snapshot are unchanged.

[mwiewior]

hey @SPPearce - could you please check if it's ready for merging now ?

[mwiewior]

Pushed 3811c74: the module now requires the reference FASTA. vepyr annotates with --everything, which needs it, so VEPYR_ANNOTATE stops with an explicit error when channel 3 carries none, instead of starting a task that vepyr rejects. --fasta is always passed, meta.yml marks fasta/fai as required, and a new test (no fasta - fails) covers it. The existing tests and snapshots are unchanged (nf-test --ci passes locally).