NVIDIA Doubles Down on AI Drug Discovery with $50M Recursion Investment

Pictured: scientist working on machine learning mo

Pictured: scientist working on machine learning mo

The tech giant is investing $50 million in the Utah-based biotech to accelerate development of its AI foundation models for drug discovery.

Pictured: a scientist working on machine learning models/gorodenkoff/iStock

Clinical-stage biotech Recursion continues to go from strength to strength. After its recent acquisition of Cyclica and Valence, the company announced a partnership with—and $50 million investment from—NVIDIA on Wednesday.

Under the agreement, the tech giant is giving Recursion a $50 million private investment in public equity. The two companies are looking to accelerate the development of Recursion’s AI foundation models to help the biopharma industry create improved patient therapies more quickly.

“With our powerful dataset and NVIDIA’s accelerated computing capabilities, we intend to create groundbreaking foundation models in biology and chemistry at a scale unlike anything that has ever been released in the biological space,” Recursion CEO Chris Gibson said in a statement.

Recursion has built a massive biological and chemical database via its Recursion OS, which exceeds 23 petabytes and has three trillion searchable compound and gene relationships. The company has been training machine-learning algorithms on this dataset to tackle the complex challenges of drug discovery.

The partnership will enable Recursion to leverage NVIDIA’s DGX Cloud to accelerate its AI training, which offers AI-as-a-service software. The deal will also give Recursion access to the NVIDIA AI-stack and its software expertise to further enhance the training of its models.

These models will also be able to access NVIDIA’s BioNeMo, a cloud service for generative AI in drug discovery, which was released earlier this year. The cloud service offers tools for customization and deployment of biomolecular models through cloud APIs, which Recursion intends to support its internal pipeline with, as well as its current and future partners. Ultimately, those models can then be licensed to other biotech and pharma companies through the cloud service.

Wednesday’s news has already paid off for Recursion shareholders. The company’s shares jumped more than 80% after the announcement. Recursion’s market value rose by nearly $1 billion, an all-time high for the company which was publicly listed in 2021.

That same year, Roche signed a potentially multi-billion dollar deal with Recursion to access the Recursion OS and Recursion Map to develop as many as 40 new medications for oncology and neuroscience.

Last month, Recursion announced that one of its drug candidates was enrolled for its Phase II SYCAMORE clinical trials ahead of schedule. The company has been testing REC-994, a small molecule superoxide scavenger and therapeutic candidate for treatment of cerebral cavernous malformation, a disease that causes malformation of the brain and spinal cord. Recursion also has four other drug candidates undergoing clinical trials at various stages.

Tech giants like Google and NVIDIA have been teaming with biopharma companies.

Earlier this year, Bayer AG and Google Cloud announced a collaboration in early drug discovery using Google Cloud’s Tensorflow Processing Units, which are custom-developed accelerators designed to run machine learning models and computationally-intensive workloads, to help accelerate and scale Bayer’s quantum chemistry calculations.

Connor Lynch is a freelance writer based in Ottawa, Canada. Reach him at lynchjourno@gmail.com.

MORE ON THIS TOPIC