Silicon Valley Giant NVIDIA Partners with Arc Institute and Stanford University to Launch Evo 2 Foundation Model for Genomic Data
Leading technology company NVIDIA has announced a groundbreaking partnership with the Arc Institute and Stanford University to introduce Evo 2, a cutting-edge foundation model that comprehends the genetic code across all domains of life. This revolutionary model, known as the largest publicly available AI model for genomic data, was developed utilizing the NVIDIA DGX Cloud platform on Amazon Web Services (AWS) in collaboration with Arc and Stanford.
NVIDIA BioNeMo Platform Accessible to Global Developers
Evo 2 is now accessible to developers worldwide through the NVIDIA BioNeMo platform, including as an NVIDIA NIM microservice for streamlined and secure AI deployment. The model has been trained on a vast dataset of nearly 9 trillion nucleotides, the fundamental building blocks of DNA and RNA. Evo 2’s capabilities extend to various biomolecular research applications, such as predicting protein form and function based on genetic sequences, identifying novel molecules for healthcare and industrial purposes, and assessing the impact of gene mutations on function.
Customizable Biological Sequences with NVIDIA NIM Microservice
The NVIDIA NIM microservice for Evo 2 empowers users to generate a diverse range of biological sequences, with adjustable model parameters. Additionally, developers interested in refining Evo 2 with proprietary datasets can download the model via the open-source NVIDIA BioNeMo Framework, a suite of accelerated computing tools tailored for biomolecular research.
Envisioning the Future of Generative Genomics
Patrick Hsu, Co-founder of the Arc Institute and Assistant Professor of Bioengineering at the University of California, Berkeley, hailed Evo 2 as a significant advancement in generative genomics. He emphasized the potential for Evo 2 to drive innovative solutions in healthcare and environmental science that were previously unimaginable.
According to Dave Burke, Chief Technology Officer at Arc, deploying a model like Evo 2 is akin to deploying a powerful new telescope to explore the furthest reaches of the universe. The possibilities for discovery are immense, and the impact on scientific exploration could be transformative.
The Arc Institute’s Commitment to Long-Term Scientific Research
The Arc Institute provides researchers with multi-year funding to tackle enduring scientific challenges, enabling scientists to focus on creative research endeavors rather than grant writing. This approach fosters innovation and breakthroughs in various fields of study, including genomics and biotechnology.
The Larger Trend in AI and Healthcare
NVIDIA’s Ongoing Innovations in Healthcare Technology
In 2022, NVIDIA expanded its healthcare product portfolio with the launch of Clara Holoscan MGX, a tool designed to assist medical device organizations in developing AI solutions that meet regulatory standards. Building on the success of Clara Holoscan, Clara Holoscan MGX offers real-time insights by processing high-throughput data streams efficiently.
Collaborative Efforts with Pharma Companies and Research Institutions
In collaboration with AstraZeneca and the University of Florida, NVIDIA developed the MegaMoIBART drug-discovery model to enhance drug discovery processes and patient care. Utilizing transformer neural networks on NVIDIA’s computational drug discovery platform, Clara Discovery, MegaMoIBART facilitates reaction prediction, molecular optimization, and de novo molecular generation.
Furthermore, NVIDIA partnered with Harvard University on AtacWorks, an AI-based toolkit that accelerates DNA analysis, enabling researchers to conduct whole genome analyses in just 30 minutes. This innovative tool identifies specific sequencing data and locates accessible DNA regions for in-depth genetic insights.
Arc Institute’s Arc Virtual Cell Atlas Initiative
This month, the Arc Institute launched the Arc Virtual Cell Atlas, a comprehensive resource containing computation-ready single-cell measurements from over 300 cells. The Atlas serves as a foundational platform for AI-driven biological discovery, offering researchers valuable insights into cellular data to drive future innovations in genomics and biotechnology.
