NVIDIA and partners add viral protein complex structures to AlphaFold Database
New viral protein complex data and open tools aim to speed treatments and vaccines for future outbreaks.
NVIDIA, Google DeepMind, EMBL-EBI and other partners have published predicted 3D structures for protein complexes from more than 2,800 human-infecting viruses in the AlphaFold Database, aiming to improve readiness for future pandemics. The dataset was generated using AlphaFold2 with NVIDIA’s BioNeMo Inference Runtime to scale predictions across thousands of viral proteomes, and about 30% of the protein interactions are structurally novel relative to the Protein Data Bank. NVIDIA is also open-sourcing the GPU-accelerated BioNeMo Structure Prediction Pipeline so researchers can run their own sequence-to-structure workflows. The release lands as AlphaFold’s database surpasses 260 million protein and protein complex predictions and is positioned as a freely accessible resource, including for scientists in low-resource settings. By focusing on viral protein complexes — the typical targets of vaccines and antivirals — the project aims to provide a structural “stockpile” that speeds up diagnostics and therapeutic development when new outbreaks emerge.
Why it matters
Publishing 3D structures for protein complexes from thousands of human-infecting viruses gives researchers a large structural reference they did not have before, including many interactions not seen in existing databases. Combined with an open-source pipeline for running their own predictions, scientists worldwide, including in low-resource settings, gain earlier insight into viral targets for diagnostics, drugs and vaccines, which can help shorten the path from an emerging outbreak to medical countermeasures.
Signal or noise?
Does this story matter, or is it hype? Decide before you see what everyone else thinks.
Sources
- NVIDIA Blog