NetGoodIndexSubmit a correction

Benefit Ledger · verified · Biology

RFdiffusion designs binders and assemblies later confirmed by experiment

The Baker lab’s RFdiffusion, published in Nature on 11 July 2023, generated protein backbones whose experimental structures and binding functions, including an influenza HA binder, matched designs.

11 Jul 2023Tier 4 HistoricMethodology 0.1

Current score

+8.45

30 base · Historic (tier 4 of 5, 30 pts)
× 0.7500 attribution · Critical contribution
× 0.8500 evidence · Multiple independent confirmations
× 0.6500 realization · Independently validated or deployed
× 0.8500 durability
Event-level product before credit split: 10.57

General, experimentally validated generative design method with wide uptake (tier 4). Attribution shared with RoseTTAFold and wet lab (0.75). Independent adoption plus cryo-EM (0.85). Realization is a working method used broadly, not approved drugs.

What happened

Fine-tuning RoseTTAFold as a diffusion model enabled unconditional monomers, binders, oligomers, and motif scaffolding. Hundreds of designs were characterized; a cryo-EM structure of an HA binder closely matched the design model. Open-source code spread quickly. Human specification of targets and experimental filtering remained essential.

Model attribution

+8.45

RoseTTAFold

Diffusion model that generated backbones later validated experimentally.

Central design engine; sequence design and experiments were additional human/ML steps.

Attribution 0.7500 · Credit share 80% · University of Washington Institute for Protein Design

Claims

  • RFdiffusion designs were experimentally characterized, including a cryo-EM-confirmed influenza HA binder.

    outcome · supported

Sources

primary sources

Secondary domains: Medicine & Health, Materials Science

Revision history

  • 13 Sep 2026 · 0.00 → 8.45

    Initial adjudicated seed score under methodology 0.1.

RFdiffusion designs binders and assemblies later confirmed by experiment · NetGoodIndex