NovaAtom-Lite hits 79.07% on FoldBench antibody prediction, far above AlphaFold 3's 47.9%
QuanquanGu · x · 2026-11-11
Geodesiclab launched NovaAtom-Lite, an all-atom structure prediction model for drug discovery that predicts structures of proteins, nucleic acids, and their bound molecules, including how they fit together.
On FoldBench antibody–antigen structure prediction (DockQ ≥ 0.23), it outperforms both IsoDDE and AlphaFold 3:
- NovaAtom-Lite: 79.07% (n=172)
- IsoDDE: 75.58%
- AlphaFold 3: 47.90% (n=167)
Evaluation uses 5 seeds × 5 samples with top-1 selected by the model's own ranking score. Available today via the Geodesic API and NovaDDE.
More from Research
- LongTail-SWE benchmark: 202 niche-language coding tasks, GPT 6.1 Sol loses to 5.6 Sol — teortaxesTex · 2026-11-11
- Sara Seager and son challenge 'water chauvinism': sulfuric acid as alien life's solvent — _onionesque · 2026-11-11
- David Bau's Cornell Tech talk on interpretability, indirection and introspection now on YouTube — yoavartzi · 2026-11-11
- Cisco releases Signals 2.0 to auto-diagnose 45 types of AI agent failures, hitting 91.43% F1 — ptkbhv · 2026-11-11
- Review of 6 RSI benchmarks finds flawed tasks in 5, with 31-63% error rates — teortaxesTex · 2026-11-11
- Freezing the VLM costs π models 23-56 points but barely hurts GR00T — rahulshivchand · 2026-11-11