TL;DR

AstraZeneca has signed a multi-year deal with California AI biotech insitro to find new treatments for MASH, a serious liver disease, and cardiomyopathies, which damage the heart muscle, the British drugmaker said on Thursday 15 October, Reuters reports. AstraZeneca pays about £29 million ($40 million) upfront. Insitro could earn up to roughly £1.07 billion ($1.46 billion) more in development and commercial milestones, plus single-digit royalties on sales.

Both conditions have long lacked effective treatments because scientists have struggled to identify what drives them biologically, AstraZeneca said.

How the work is split

Insitro will apply its “Virtual Human” platform, which pairs cell models grown in the lab with machine learning, to find causal disease biology. It will take targets that AstraZeneca nominates from discovery through to development-candidate stage. AstraZeneca then takes over, running and paying for any clinical trials.

Sharon Barr, who heads biopharmaceuticals R&D at AstraZeneca as an executive vice president, told Reuters that choosing targets is among the hardest calls in research. By the time a new pathway appears in the major journals, she said, “it is no longer novel”, while spotting such connections early and testing them in cells puts the company “ahead of the game”. She added that combining AI with experienced scientists means fewer molecules need testing in the lab.

Part of a pattern for insitro

The deal adds to insitro’s growing list of pharma partners. It recently widened its work with Bristol Myers Squibb on targets for amyotrophic lateral sclerosis (ALS), and has a separate project with Eli Lilly building models that predict how small molecules behave in living organisms.

Looking forward

Most of the potential value sits in the milestones, not the upfront fee. In our view, the interesting part for UK life sciences is where the AI sits. AstraZeneca is buying help with the earliest and most uncertain step, picking the target, while keeping the clinical stages in-house. If the approach works, it could encourage other large drugmakers to license AI platforms for target discovery rather than build their own.