How Open Science Can Help Researchers Prepare for the Next Pandemic
The predicted structures were generated using AlphaFold2 optimized with NVIDIA BioNeMo Inference Runtime and added to the AlphaFold Database.
- About 30% of the viral protein interactions in the dataset have never been documented in the Protein Data Bank
- NVIDIA also open-sourced the BioNeMo Structure Prediction Pipeline, the GPU-accelerated workflow used to generate the dataset
- The dataset covers viral families known to infect humans, ranging from common-cold viruses to emerging threats like Mpox
Virologists and drug developers gain open structural models for thousands of uncharacterized viral protein complexes to accelerate vaccine and treatment design.
