New module: GRIMER

[Jonahnki]

Closes #11566

Description

Adds a new nf-core module for GRIMER (v1.1.0), a tool that generates interactive HTML dashboards for contamination detection in metagenomics and viromics datasets.

Developed during the May 2026 nf-core x VirJenDB Virus Bioinformatics Hackathon.

Checklist

• environment.yml matches bioconda package
• main.nf passes lint (47/47 tests passed)
• meta.yml complete and schema-valid
• nf-test dry run passes (2/2 tests discovered)
• All pre-commit hooks pass

[Jonahnki]

The stub test now uses an existing nf-core/test-datasets file to pass CI. I hope to add in a proper GRIMER-specific count table TSV to nf-core/test-datasets in a follow-up post-PR after review

[SPPearce]

Can you create this file on-the-fly within the nf.test please.

[SPPearce]

Call this config or something similar

[Jonahnki]

Thanks for the review @SPPearce! You're definitely right with the meta3 / config naming. I will rename to config since it's a pipeline-level file not tied to individual samples, and restructure as a plain path input without a meta map wrapper

[SPPearce]

We use meta, meta2, meta3 to mean Nextflow meta maps, not csv file metadata.
Are these files actually sample dependent? Because if so it should be part of the initial tuple.

[Jonahnki]

Thanks for the review and comments, I'll address them

[SPPearce]

Not sure what is going on with this file, but I don't think you should be changing it.

[Jonahnki]

This file was actually modified automatically by nf-core modules lint when it prompted for repository type on first run. I Will revert it to its original state and exclude it from the PR

[SPPearce]

This should be the snapshot of the outputs.

[Jonahnki]

Thanks for the comment, I'll replace { assert process.out.report } with { assert snapshot(process.out).match() } and produce another output snapshot

[Jonahnki]

Thanks for the latest comments @SPPearce! I've gone through everything and pushed some fixe

*The metadata input in main.nf was tripping the linter since anything prefixed with meta gets treated as a Groovy map and renamed it to sample_metadata and that cleared it up
*Synced meta.yml to match while I was at it; --fix had silently collapsed the sample_metadata block to {} so I restored the full field descriptions manually. All 51 lint checks are green now
*For the test part, I swaped out the bare assert calls for snapshot(process.out).match() and added a -stub block. Test data is now generated on-the-fly in setup {} so the static tests/testdata/test_count_table.tsv wasn't needed anymore and I've removed it
*.nf-core.yml is also back to upstream state

[Jonahnki]

Managed to fix the two GRIMER-specific CI failures in the latest push:

*Version eval: grimer --version outputs a multi-line ASCII art banner rather than a clean version string, which caused the snapshot mismatch on CI. Changed the eval to grimer --version 2>&1 | tail -1 which reliably extracts just the bare version number from the last line of the output
*meta.yml sync: ran nf-core modules lint --fix to update the version eval string in both the outputs and topics blocks to match the updated main.nf. All 51 lint checks now passing and the stub snapshot has been regenerated to match

The remaining CI failures (SENTIEON, HMMER, WFMASH, SALSA2, KRAKENUNIQ etc.) are infrastructure-level flakes — runner disk full (no space left on device) and conda HTTP 403 errors and unrelated to this PR

[Jonahnki]

The confirm-pass-lint gate failure is downstream of the multiqc and multiqcsav lint failures, which seems like pre-existing issues in those modules unrelated to this PR. The GRIMER module lint passes cleanly. I'm happy to have a reviewer confirm it