| name | alphafold2 |
| description | Predict or audit protein structures with AlphaFold2-style workflows. Use when a research task needs monomer/multimer structure prediction, MSA/template handling, confidence metrics, or comparison against PDB/AlphaFold references. |
AlphaFold2
Use this skill for protein-structure prediction or parity checks around AlphaFold2-style outputs.
Workflow:
- Capture the biological question, sequence identifiers, FASTA input, oligomer state, organism, and expected cofactors or partners.
- Verify the execution route before running: local install, managed endpoint, Modal/SSH job, or a documented public source. Do not assume weights, databases, or GPUs exist.
- Save inputs, command or endpoint payload, model settings, stdout/stderr, and raw outputs under the active Feynman artifact folder.
- Report pLDDT, PAE, ranking confidence, chain coverage, truncation, templates/MSA provenance, and any residues or interfaces that should not be trusted.
- Compare against known structures or AlphaFold DB records when the claim depends on novelty, domain movement, interface geometry, or mutation impact.
Outputs should include the FASTA, predicted PDB/mmCIF, confidence files when available, a short method note, and a .provenance.md sidecar.
