Nvidia invests $50 million in Recursion to train AI models for drug discovery
Views: 1738
2023-07-12 20:52
(Reuters) -Nvidia invested $50 million to speed up training of biotech firm Recursion's artificial intelligence models for drug discovery, which

(Reuters) -Nvidia invested $50 million to speed up training of biotech firm Recursion's artificial intelligence models for drug discovery, which the chip designer can then potentially license out, the companies said on Wednesday.

Recursion shares jumped about 167% in premarket trading on the news.

Recursion will train AI models on Nvidia's cloud platform using its own biological and chemical datasets exceeding 23,000 terabytes.

Nvidia could then license the AI models to biotech firms via BioNeMo, a generative AI cloud service for drug discovery that it rolled out widely earlier this year.

The companies did not specify whether Nvidia will take a stake in Recursion. The drug developer did not immediately respond to a Reuters request for clarification.

Recursion, which has partnered with drugmakers such as Bayer and Roche, plans to use Nvidia's software to support its own pipeline as well as current and future partners.

The biopharma company, which went public in 2021, runs the Recursion OS platform that gives drugmakers access to datasets required for designing and developing therapeutics.

Recursion is conducting human trials for five of its drugs, including a candidate that is in mid-stage trials to treat a type of neurovascular disease caused by malformation of small blood vessels in the brain. The company uses AI to discover drugs.

Recursion's other backers include Abu Dhabi sovereign investor Mubadala and UK-based investment firm Baillie Gifford & Co.

In May, Recursion, whose advisers include AI pioneer Yoshua Bengio, had agreed to buy two companies in the AI-driven drug discovery space for $87.5 million.

(Reporting by Chavi Mehta in Bengaluru and Stephen Nellis in San Francisco; Additional reporting by Mariam Sunny in Bengaluru; Editing by Shilpi Majumdar)

Tags recursion pharms nvidia epus smart epus scitech