NVIDIA, DeepMind open AlphaFold-predicted structures for 2,800+ viral protein complexes
NVIDIA Blog · rss · 2026-09-24
NVIDIA has joined a coalition including Google DeepMind and EMBL-EBI to openly release predicted 3D structures for protein complexes of more than 2,800 viruses via the AlphaFold Database, aiming to stockpile knowledge ahead of the next pandemic.
- Structures were inferred with AlphaFold2, optimized via NVIDIA BioNeMo Inference Runtime to scale across thousands of viral proteomes; roughly 30% of the protein interactions are entirely new to science.
- Traditional X-ray crystallography takes years and thousands of dollars per structure; AlphaFold2 predicts in minutes, letting the team systematically cover viral families known to infect humans, from common-cold viruses to Mpox.
- NVIDIA is also open-sourcing the GPU-accelerated BioNeMo Structure Prediction Pipeline used to generate the dataset, so researchers can predict structures for their own targets.
- The AlphaFold Database now holds 260M+ protein and complex predictions. Citing an estimate of a 50% chance of a COVID-scale pandemic by 2050, collaborators frame the dataset as an engine for hypothesis generation in vaccine and drug design.
More from Research
- Anthropic: ~950 Claude Agents Spent 21 Hours to Find an Unknown Enzyme System — VraserX · 2026-09-25
- Neural Spectral Capacity: a training-free architecture score that beats #Params and prunes LLaMA-7B 5900x faster — CityU-HongKong · 2026-09-25
- OpenAI reportedly spent $10M in a week on a Navier-Stokes counterexample — funding 50 mathematicians for a year — RexDouglass · 2026-09-25
- Economists call for zero tolerance on AI-written papers: reject, retract, and ban authors — paulnovosad · 2026-09-25
- HRI 2027 Adds Archival Industry White Paper Track for Real-World Robot Deployments — petitegeek · 2026-09-25
- Study: Reasoning hurts 15.7% of multimodal embeddings; training-free SURE router fixes it — _reachsumit · 2026-09-25