AlphaFold

5-30 minutesPaperCodeNeed >20GB

Accurate and quick protein structure prediction for multimers and monomers

Readiness

Readiness: Need >20GB
Smoke: passed
Artifacts: passed
Production: need_gt20gb_colabfold_local_databases_or_precomputed_msa_strategy
Storage: >20GB; full ColabFold local databases documented around 940GB
Strategy: Use ColabFold container/SIF plus either full local MMseqs/ColabFold databases on larger storage or a precomputed-MSA/server-MSA production policy.
Next: Provision large database storage or approve a server-MSA/precomputed-MSA scope; then install ColabFold SIF, download databases or model weights as needed, wire real adapter, and run Slurm/API benchmark.

2026-07-18T17:33:47+08:00 SOP large-tool batch: reclassified to Need >20GB. Official ColabFold docs state local database setup needs a disk with about 940GB. Backend currently has about 61GB free, so full local AlphaFold/ColabFold database production cannot be installed on current storage. Existing runner remains alphafold-smoke placeholder; no real ColabFold SIF/database/API Slurm validation exists.