Replace BioMart MartsAdapter with epytope 4.0.0 EnsemblRESTAdapter

[jonasscheid]

Description

Drop in epytope 4.0.0's EnsemblRESTAdapter as a replacement for the BioMart MartsAdapter in bin/epaa.py, to eliminate the intermittent BioMart lookup failures in variant prediction. Validated end-to-end on GRCh37 and GRCh38 test VCFs: peptides, FASTA and all Ensembl-based annotations are identical to the MartsAdapter output (only the supplementary refseq column returns a subset of cross-references).

• Swap MartsAdapter → EnsemblRESTAdapter in epaa.py; GENOME_REFERENCE_MAP now points at Ensembl REST servers.
• Bump epytope 3.3.1 → 4.0.0 in the EPYTOPE_VARIANT_PREDICTION container and conda env.
• Fix pandas 3.0 positional Series access (pep[0] → pep.iloc[0]) pulled in by epytope 4.0.0.

Draft / pending: epytope 4.0.0 is not yet on bioconda (no recipe bump PR exists yet), so the container tag and build hash are placeholders. This PR is blocked until epytope 4.0.0 is merged into bioconda and the biocontainer is built.

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 pipeline conventions in the contribution docs
• If necessary, also make a PR on the nf-core/epitopeprediction branch on the nf-core/test-datasets repository.
• Make sure your code lints (nf-core pipelines lint).
• Ensure the test suite passes (nextflow run . -profile test,docker --outdir <OUTDIR>).
• Check for unexpected warnings in debug mode (nextflow run . -profile debug,test,docker --outdir <OUTDIR>).
• Usage Documentation in docs/usage.md is updated.
• Output Documentation in docs/output.md is updated.
• CHANGELOG.md is updated.
• README.md is updated (including new tool citations and authors/contributors).

[jonasscheid]

Known limitation — mouse GRCm38/mm10

Previously GENOME_REFERENCE_MAP routed grcm38/mm10 to the nov2020.archive.ensembl.org BioMart (the last release carrying GRCm38). The Ensembl REST API has no per-assembly archive server — rest.ensembl.org only serves the current mouse assembly (GRCm39) — so with EnsemblRESTAdapter these references now resolve against GRCm39.

In practice this means: transcript IDs that still exist in GRCm39 resolve fine, but any transcript retired between GRCm38 and GRCm39 will fail to look up. Human GRCh37 is unaffected, since Ensembl does provide grch37.rest.ensembl.org.

Flagging for discussion before this comes out of draft — options are to (a) accept GRCm39-only mouse support, (b) keep a BioMart fallback for the archive-only mouse builds, or (c) drop the grcm38/mm10 entries to avoid implying support we can't deliver via REST.