Technology from computing powerhouse Nvidia is already hard at work inside companies at the convergence of technology and biology, including the supercomputer driving Recursion Pharmaceutical’s drug discovery efforts. The companies are getting a little cozier. Nvidia is investing $50 million in Recursion. More than working capital, the deal marks the start of a partnership.
Recursion uses technology to interrogate data, gaining insights into biological and chemical relationships. The Salt Lake City-based company says its work to date has built up a dataset that exceeds 23 petabytes and 3 trillion searchable relationships between genes and compounds. The company applies its tech-driven approach to internal drug discovery research as well as to partnerships it has with Roche and Bayer.
A Personalized Approach to Medication Nonadherence
At the Abarca Forward conference earlier this year, George Van Antwerp, managing director at Deloitte, discussed how social determinants of health and a personalized member experience can improve medication adherence and health outcomes.
Going forward, Recursion will collaborate with Nvidia to train foundation models, which are artificial intelligence programs that analyze massive amounts of data to find patterns and make predictions. Taran Loper, Recursion’s director of communications, said in an email that the company is considering releasing some of its machine learning and artificial intelligence models to commercial partners through BioNeMo, Nvidia’s cloud-based platform for generative AI drug discovery. Recursion anticipates using BioNeMo to support its internal pipeline and the research efforts of its current and future partners.
The Recursion supercomputer, BioHive-1, uses graphics processing units (GPUs) from Nvidia. Loper said Nvidia will provide recursion with priority access to its latest GPUs, offering engineering support to optimize the company’s models. There’s also potential for the companies to join in research efforts, he added.
“Despite the massive computational horsepower we have built in-house with BioHive-1, training some of the most advanced AI foundation models in the world across large subsets of our 23 petabyte proprietary dataset means we need even more compute [‘compute’ is industry shorthand for ‘computing power’] to go as fast as possible,” Loper said. “This partnership provides us access to that compute alongside Nvidia’s industry-leading expertise, resources, and shared vision, which we hope will unlock new possibilities and accelerate the development of much-needed medicines.”
Recursion added to its technology toolbox in May through the acquisitions of two AI-startups that help the company develop new chemistry capabilities. Together, the stock transactions amounted to about $88 million.
BioLabs Pegasus Park Cultivates Life Science Ecosystem
Gabby Everett, the site director for BioLabs Pegasus Park, offered a tour of the space and shared some examples of why early-stage life science companies should choose North Texas.
Nvidia’s investment in Recursion comes in the form of a stock purchase. The technology giant is buying 7.7 million shares of Recursion for $6.48 apiece, according to a regulatory filing. Shares of Recursion more than doubled after the deal was announced Wednesday.
Photo: metamorworks, Getty Images
