DFMDock

5-30 minutesPaperCodeNeed >5GB

Protein-protein docking using diffusion and flow matching

Readiness

Readiness: Need >5GB
Smoke: not_tested
Artifacts: not_applicable
Production: need_gt5gb_cuda_pyg_esm_docking_runtime_pending
Storage: 5-20GB
Strategy: dedicated CUDA 11.8 PyTorch/PyG/ESM DFMDock SIF + checkpoint/fixture overlay
Next: Build/stage dedicated DFMDock CUDA 11.8 SIF with bundled checkpoints/weights and DB5 tiny fixture, run direct Slurm src/inference_single.py receptor.pdb ligand.pdb, then wire runner.

No obvious large model/database dependency identified; validate upstream docs before install. 2026-07-05 SOP batch recheck: backend inventory found no staged DFMDock source/checkpoints/fixtures/SIF/artifacts/runner. Upstream uses Python 3.10, torch 2.4.1 cu118, torch-geometric, fair-esm, Lightning, CUDA libs, and inference_single.py. Reclassified from Smoke Pending to Need >5GB.