It generated protein backbones with specified structural features, using a diffusion model built from RoseTTAFold. A separate sequence-design model could then propose sequences for those structures.
Variant labels are model predictions, not clinical diagnoses. Meteorologist rankings compare forecasts in a particular evaluation. Protein generation supplies candidates whose function must be established through appropriate experiments.
Evidence you can use
Historical snapshot: October 2023. Dates and populations are specified per row.
| Measure | Reported value | Definition and source |
|---|---|---|
| AlphaMissense predictions | 71M missense variants | Predicted effects across the human proteome, not 71 million clinically confirmed classifications.2023 report, slide 62 (PDF page 62) |
| NowcastNet expert evaluators | 62 meteorologists | Professional meteorologists in China who evaluated the reported forecasts.2023 report, slide 57 (PDF page 57) |
| NowcastNet first-place ranking | 71% of cases | Share of evaluated cases ranked first among compared methods; not forecast accuracy across all weather.2023 report, slide 57 (PDF page 57) |
Variant labels are model predictions, not clinical diagnoses. Meteorologist rankings compare forecasts in a particular evaluation. Protein generation supplies candidates whose function must be established through appropriate experiments.
Historical snapshot published October 12, 2023. This web edition was prepared on October 10, 2026 from the online deck and original launch posts. Findings and forecasts retain their original time frame.
Benaich, Nathan. “What did RFdiffusion enable?.” State of AI Report 2023. Historical report snapshot; web edition prepared 2026-10-10.