EnzymeCoT - Stage III joint checkpoint
Joint adapters for reaction-conditioned, three-round de novo enzyme design: Qwen3-4B as the understanding model and RFdiffusion3 as the generator, trained together.
What this file is
s3_best.pt - a 553 MB PyTorch state_dict (755 tensors, 138.9 M parameters).
Adapters only: no optimizer state, no base model weights.
| Training | 4000 steps, run to completion |
| Validation | L_gen 0.386 -> 0.3497 (the lowest value over the whole run) |
| Trained parts | Qwen LoRA + structure/ligand projections + injection (2.29 M) + RFdiffusion3 LoRA (2.89 M, MLP) |
| Date | 2026-09-11 |
Important caveat about the reported numbers
Evaluation figures reported for this line of work - catalytic-site RMSD 0.94 A, LigandMPNN-8 self-consistency 5/8 and 8/8 designable, pLDDT 77.9 / 94.5 - were produced with the step-2500 checkpoint (val 0.3526).
This file is step-4000 (val 0.3497). They are not the same checkpoint, and the step-4000 weights have not been re-evaluated end to end. Please do not attribute those numbers to this file without re-running the evaluation yourself.
Requirements
This is an adapter checkpoint, not a standalone model. You need:
- Qwen3-4B base weights
- RFdiffusion3
Task definition
Given a reaction (SMILES), the ligand's chemistry and 3D coordinates, and a target protein length, the model designs an enzyme in three rounds:
- R1 - an unindexed catalytic-residue set (identity + catalytic/support role) together with local geometry relative to the ligand. Object IDs carry no sequence position.
- R2 - a full backbone, then an assignment of the round-1 objects onto positions in that generated backbone.
- R3 - the remaining sequence and side chains, inheriting the committed core identities.
Data
Trained on the EnzymeCoT corpus (19,888 train / 1,123 validation records, lengths 100-997) with a reaction- and homology-disjoint split. Ligand binding conformations are taken from reference complexes, which are Boltz predictions rather than experimental structures.
Limitations
- No experimental validation. No catalytic activity has been measured.
- The metrics above are computational self-consistency and RMSD-to-reference; neither establishes catalysis.
- Autonomous three-round design quality remains weak; treat this checkpoint as a research artifact, not a working enzyme designer.