EnzyGen2

1-5 minutesPaperCodeNeed >5GB

Motif-conditioned functional enzyme sequence and structure co-design

Readiness

Readiness: Need >5GB
Smoke: not_tested
Artifacts: not_applicable
Production: need_gt5gb_cuda_fairseq_esm_checkpoint_pending
Storage: 5-20GB
Strategy: shared CUDA PyTorch/fairseq/ESM SIF + model/cache overlay
Next: Build/stage shared CUDA 11.6/PyTorch 1.13/fairseq/ESM SIF or overlay, stage EnzyGen2 checkpoint plus ncbi2id.json and minimal example, run direct Slurm inference smoke, then wire 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 EnzyGen2 source/checkpoint/ncbi2id data/SIF/artifacts/runner. Upstream requires CUDA/PyTorch/fairseq/ESM stack plus EnzyGen2 checkpoint and mapping files, which exceeds the slim shared CPU path. Reclassified from Smoke Pending to Need >5GB.