BioEmu2 matches SOTA folding models on Runs N' Poses and FoldBench
FrankNoeBerlin · x · 2026-10-20
Structure prediction is the foundation of BioEmu2. With the same train/test split, its structure prediction mode is competitive with SOTA folding models:
- Runs N' Poses high-res protein-ligand benchmark: success rates on par with Protenix v1/v2, ESMFold2, OpenFold3, and OpenDDE.
- FoldBench: protein, nucleic acid, and complex lDDT/success rates comparable to Boltz v1, ESMFold2, and Protenix v2.
Vision: sequencing and structure prediction are at scale, but how do biomolecules change shape, bind, and drive function and disease? The team aims to build quantitative methods that scale to the genome.
More from Research
- Causal inference tutorial series covers DAG engines for separating correlation from causation — Nasereliver · 2026-10-20
- Google's FireSat constellation uses AI to spot wildfires as small as 5x5 meters from space — GoogleAI · 2026-10-20
- CyberArena benchmarks AI agents in live attack-and-defense combat — 2077AIDataFoundation · 2026-10-20
- Training MOSS-Image LoRAs locally on 8GB VRAM: 500 steps beats 8000, full settings inside — AcademiaSD · 2026-10-20
- PNAS study: brief AI chats drive 17,950 people to sign petitions and donate, attitudes don't predict behavior — _FelixSimon_ · 2026-10-20
- Citrini Used Anthropic's Internal Model to Design Scorpion Antivenom Proteins Overnight — citrini · 2026-10-20