How Open Science Can Help Researchers Prepare for the Next Pandemic
nvidia ·
NVIDIA has partnered with Google DeepMind and the European Molecular Biology Laboratory’s European Bioinformatics Institute to release an open dataset containing predicted 3D structures for viral protein complexes from over 2,800 viruses. This initiative aims to prepare scientists for future pandemics by providing crucial structural data that was unavailable during previous outbreaks. The predictions were generated using AlphaFold2, optimized with NVIDIA BioNeMo Inference Runtime, allowing for rapid scaling across thousands of viral proteomes. Approximately 30% of the included protein interactions are entirely new to science, offering novel insights for vaccine and drug development. By making this data freely accessible via the AlphaFold Database, researchers can bypass time-consuming traditional crystallization methods and accelerate hypothesis generation. The collaboration also involves releasing the GPU-accelerated BioNeMo Structure Prediction Pipeline so other scientists can predict structures for their own targets. This effort supports global pandemic preparedness, particularly aiding low-resource settings and addressing the estimated 50% chance of a severe pandemic occurring by 2050.
nvidia