NVIDIA and Google DeepMind release 2,800 viral protein structures

This collaboration aims to prepare for future pandemics by stockpiling structural knowledge ahead of time. An analysis by the Center for Global Development estimates a roughly 50% chance of a pandemic as severe as COVID-19 occurring by 2050. Joe Grove, a professor at the University of Glasgow, noted that the next pandemic may involve a novel virus lacking prior research, making this pre-emptive data crucial. The dataset includes about 30% of protein interactions that are completely new to science, offering insights not previously documented in the Protein Data Bank.

NVIDIA is also openly releasing the BioNeMo Structure Prediction Pipeline, a GPU-accelerated workflow that allows researchers to predict 3D structures for their own targets. Traditional methods for determining these structures can take years and cost thousands of dollars per structure, whereas the AI-based method predicts structures in minutes. The release coincides with a United Nations General Assembly meeting on pandemic prevention in New York City, contributing to a database that now holds more than 260 million protein predictions.