NVIDIA has unveiled a powerful new foundation model, Evo 2, which can analyze genomic data across all domains of life. Developed in collaboration with the Arc Institute and Stanford University, Evo 2 is now available to global developers on NVIDIA’s BioNeMo platform.
Evo 2 was trained on an enormous dataset of nearly 9 trillion nucleotides and can be applied to various biomolecular research applications, including predicting protein structure and function based on genetic sequences, identifying novel molecules for healthcare and industrial applications, and evaluating the impact of gene mutations on their function.
The model’s developer, Patrick Hsu, described Evo 2 as a “major milestone” for generative genomics. By advancing our understanding of genetic code, researchers can pursue solutions in healthcare, environmental science, and other fields that are currently unimaginable.
Evo 2 is accessible to developers through the open-source NVIDIA BioNeMo Framework, which provides accelerated computing tools for biomolecular research. The model’s capabilities include generating biological sequences with adjustable settings, enabling users to design complex systems more efficiently.
The Arc Institute, a nonprofit biomedical research organization, has established itself as a leader in innovative research and collaboration with AI experts like NVIDIA. Its researchers receive state-of-the-art lab space and funding for long-term projects, focusing on disease areas such as cancer, immune dysfunction, and neurodegeneration.
With Evo 2, scientists can analyze larger datasets, make more complex discoveries, and achieve results faster. The model’s potential applications span various scientific fields, including healthcare, agriculture, and materials science. As Arc’s chief technology officer Dave Burke noted, “Deploying a model like Evo 2 is like sending a powerful new telescope out to the farthest reaches of the universe – we don’t yet know what discoveries await us.”
Source: https://blogs.nvidia.com/blog/evo-2-biomolecular-ai