NVIDIA and Google DeepMind Release Predicted 3D Structures for 2,800 Viruses to Aid Pandemic Preparedness

NVIDIA, Google DeepMind, and the European Molecular Biology Laboratory’s European Bioinformatics Institute (EMBL-EBI) have released predicted 3D structures for the protein complexes of more than 2,800 viruses. This open dataset, available via the AlphaFold Database, aims to provide scientists with a head start for the next pandemic, addressing the risk that future outbreaks may involve viruses with less prior research than SARS-CoV-2.

Accelerating Structural Prediction

The structures were inferred using AlphaFold2, optimized with the NVIDIA BioNeMo Inference Runtime, allowing the team to scale predictions to thousands of viral proteomes. Unlike traditional methods like X-ray crystallography, which can take years and cost thousands of dollars per structure, this AI-driven approach predicts structures in minutes. NVIDIA is also openly releasing the BioNeMo Structure Prediction Pipeline, enabling researchers to generate 3D structures for their own protein targets.

New Scientific Insights

Approximately 30% of the protein interactions in the dataset are completely new to science, documenting shapes never before recorded in the Protein Data Bank. The project systematically covered viral families known to infect humans, ranging from common-cold viruses to emerging threats like Mpox. These predictions are labeled by confidence levels and are intended to serve as an engine for hypothesis generation, helping biologists understand how proteins interact within viral complexes rather than as isolated molecules.

Global Collaboration and Access

The initiative involves a coalition including the Coalition for Epidemic Preparedness Innovations, Seoul National University, Sungkyunkwan University, the Swiss Institute of Bioinformatics, and the University of Glasgow. The release coincides with a United Nations General Assembly meeting on pandemic prevention in New York City. By making this data open, the collaboration seeks to lower barriers for scientists in low-resource settings and accelerate the development of diagnostics, treatments, and vaccines.