AstraZeneca strikes $40 million AI drug discovery deal with insitro

By Patrick Wingrove

LONDON, Oct 15 (Reuters) - AstraZeneca AZN.L has teamed up with artificial intelligence biotech company insitro to discover and develop new treatments for difficult to treat liver and heart conditions, the British drugmaker said on Thursday.

Under the agreement, AstraZeneca will pay $40 million up front, with California-based insitro eligible for up to about $1.46 billion in potential development and commercial milestones, as well as a single-digit percentage of future sales in royalties.

The multi-year collaboration will focus on metabolic dysfunction-associated steatohepatitis (MASH), a severe liver disease, and cardiomyopathies, a group of diseases that affect the heart muscle. Both have historically lacked effective therapies because researchers have struggled to pinpoint their root biological drivers, the drugmaker said.

Insitro will use its "Virtual Human" AI platform, which combines lab-grown cell models and machine learning, to identify causal disease biology and advance AstraZeneca-nominated targets from discovery to development-candidate stage. AstraZeneca will then run and fund future clinical trials.

The deal builds on insitro's expansion across the pharmaceuticals industry. The biotech company recently expanded its partnership with Bristol Myers Squibb BMY.N to advance a broader portfolio of targets for amyotrophic lateral sclerosis (ALS). It also teamed up with Eli Lilly LLY.N to build AI models that predict pharmacological properties of small molecules in living organisms.

"Choosing the new targets that we want to put into our discovery pipeline is one of the hardest things we do across research and development," Sharon Barr, executive vice president of biopharmaceuticals R&D at AstraZeneca, told Reuters.

"By the time we read about a new pathway or target (in major scientific journals), it is no longer novel. It's in the public domain and we're already a few steps too late. But when we identify these unique connections early and test them in cellular systems, we're ahead of the game."

Barr added that pairing AI tools with experienced scientists allows the company to test fewer molecules in the lab, reduce guesswork and speed the search for effective medicines.