← Back to Home

Finding New Drug Targets Through Disease Differences: AstraZeneca and insitro Sign AI Research Collaboration

A $40 million upfront payment supports drug discovery starting with human and cellular data. The collaboration targets MASH and cardiomyopathy, but whether disease clues identified by AI can translate into effective therapies still requires experimental and clinical validation.

By SURL BioNews

A single diagnosis may encompass different causes of disease. For drug developers, identifying which biological mechanisms actually drive disease progression is key to selecting therapeutic targets. AstraZeneca and insitro announced a collaboration on October 15 to use artificial intelligence to analyze human and cellular data, seeking new targets and candidate therapies for metabolic dysfunction-associated steatohepatitis (MASH) and cardiomyopathy.

The globally exclusive research collaboration includes a $40 million upfront payment. According to announcements from insitro and its legal adviser Goodwin, insitro is also eligible to receive up to approximately $1.46 billion in milestone payments covering drug discovery, development, and commercialization, along with single-digit percentage royalties on future global sales. Most of the potential consideration depends on subsequent results and is not revenue available upon signing.

At the core of the collaboration is insitro’s Virtual Human platform. According to the company, the platform combines large-scale human and cellular data, using genetic information to identify biological mechanisms that may drive disease. In this collaboration, AI’s role is to help identify targets worth intervening in, then support the discovery and design of drug candidates.

The two companies see differences among patients with MASH and cardiomyopathy as a starting point for this approach. Patients within the same disease category may have different biological characteristics; the research aims to identify more specific clues to disease causation from these differences. However, whether associations in the data represent causal mechanisms that drugs can modify still requires validation through subsequent experiments.

The partners have also defined their research and development responsibilities: insitro will use its platform to explore disease mechanisms and advance targets designated by AstraZeneca to the development candidate stage, while AstraZeneca will handle clinical development, regulatory affairs, and commercialization. AstraZeneca also said it will obtain globally exclusive rights to advance selected programs under the collaboration.

The announcements to date have not disclosed the names of the collaboration’s targets, the number of programs, or the size and composition of the datasets used. Nor have they provided model performance, experimental validation, or human efficacy data for this collaboration. The deal can therefore be viewed as a pharmaceutical company’s investment in a research and development approach; it cannot yet serve as evidence that the platform has improved drug development success rates or identified effective therapies.

The next substantive test is whether the clues proposed by AI can withstand biological experiments and yield drug candidates suitable for human trials. Even if target selection is supported, safety, effective doses, and patient benefit must still be established step by step. This collaboration first opens a path for exploration, with multiple hurdles remaining before therapeutic outcomes can be achieved.

References

  1. insitro
  2. Goodwin