Assemble modelCIF - DUMMY implementation without touching workflow wiring

[keiran-rowell-unsw]

Part implements an ASSEMBLE_MODELCIF{} process as sketched out in this project board.

• DUMMY files used to populate ModelCIF fields
  • This PR avoids touching upstream workflow wiring intentionally: to merge first as test-suite then wire in values from pipeline carefully
  • DUMMY_METRIC.tsv now uses 'realistic' 20 residues values, and is synced with docs/output.md
• Comprehensive nf-test suite:
  • Uses the RCSB CifCheck program to validate against the ModelArchive dictionary
  • Reject completely invalid .mmcif; distinguishes from .pdb; and uses a real ma-.cif deposition to validate fields
  • assemble_modelcif_X has ext.args config files to test creation of .bcif, or PAE embedded or linked
• msa_tool used to specify .protocol.CoevolutionMSAStep() since the steps in-container aren't always inspectable
• DUMMY_SOFTWARE_DETAILS used to handle minimal protocol ingest into .mmcif classes for now
• populate_modelcif.py has a variety of _helper() local functions leading to a build_modelcif() function populated by argparse

DATABASES: databases will not be handled in .data.Datagroup.ReferenceDatabasein this PR. populate_modelcif.py is getting quite long already. Plus, it's a separate concept that can tie into the work done for reference dataset at NCI.

#575 might make this database handling easier, if considered valuable

DRAFT: still in draft as I'm LLM'ing and doc'ing through features and will got back for deeper inspection of .mmcif spec when ready to review

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!
• Make sure your code lints (nf-core pipelines lint).
• Check for unexpected warnings in debug mode (nextflow run . -profile debug,test,docker --outdir <OUTDIR>).
• CHANGELOG.md is updated.

[jscgh]

Tom and I were thinking that, since they're used for tests, the DUMMY files in assets/ might be best uploaded to the testdata repo (https://github.com/nf-core/test-datasets/tree/proteinfold/testdata) with the other test files.

[keiran-rowell-unsw]

Tom and I were thinking that, since they're used for tests, the DUMMY files in assets/ might be best uploaded to the testdata repo (https://github.com/nf-core/test-datasets/tree/proteinfold/testdata) with the other test files.

Hadn't thought of that, makes a lot of sense! I put them in assets/ for a faster dev loop, sounds good when it's ready!