MULTI-evolve

5-30 minutesPaperCodeNeed >5GB

Model-guided directed evolution workflow for multi-mutant nomination and assembly design

Readiness

Readiness: Need >5GB
Smoke: not_tested
Artifacts: not_applicable
Production: need_gt5gb_cuda_pytorch_pyg_plm_runtime_pending
Storage: 5-20GB
Strategy: shared CUDA PyTorch/PyG/PLM SIF + offline example-data overlay
Next: Move to 5-20GB shared CUDA/PyTorch/PyG/PLM SIF round, stage source and example data, decide W&B/offline policy, run direct Slurm example workflow, then add runner adapter.

No obvious large model/database dependency identified; validate upstream docs before install. 2026-07-05 SOP batch recheck: backend inventory found no staged MULTI-evolve source/data/SIF/artifacts/runner. Upstream requires Python 3.11, PyTorch 2.6 CUDA 12.4, PyG extensions, fair-esm, ankh, and optional PLM zero-shot path; this is a shared CUDA/PyG/PLM runtime phase. Reclassified from Smoke Pending to Need >5GB.