AI Proteins and Astellas Sign $400 Million Drug Discovery Deal

AI Proteins has signed a research collaboration and licensing agreement with Astellas Pharma to discover and develop new miniprotein-based medicines using artificial intelligence (AI). The deal could be worth more than $400 million in upfront and milestone payments, with AI Proteins also eligible to receive royalties from sales of any resulting medicines.

Announced on October 21, 2026, the partnership will combine AI Proteins’ computer-based protein design technology with Astellas’ drug development and commercialisation expertise. The companies aim to create new therapeutic candidates against multiple targets that have not yet been publicly disclosed.

Under the agreement, AI Proteins will use its proprietary AI-driven platform to design and improve miniproteins for potential use as medicines. These proteins will be created from scratch rather than adapted from proteins found in nature.

Miniproteins are small proteins that can be designed to bind to specific targets in the body. Researchers are exploring their potential to treat diseases because their size and design may offer advantages in targeting, stability and delivery. However, the potential benefits of the candidates developed through this collaboration will need to be established through further research and clinical testing.

AI Proteins uses a process known as de novo protein design, which involves building new proteins from the ground up. By combining computational design with laboratory testing, the company aims to develop molecules with specific properties suited to their intended medical use.

The approach differs from traditional drug discovery, which often begins with naturally occurring molecules or modifies existing biological structures. AI Proteins believes designing proteins for a defined purpose could help researchers explore new treatment options that may be difficult to achieve with conventional biologic medicines.

Under the terms of the agreement, Astellas will have exclusive worldwide rights to therapeutic candidates developed through the collaboration. Once the candidates are ready to move forward, Astellas will take responsibility for further development, regulatory submissions and commercialisation.

AI Proteins will receive an upfront cash payment for designing and optimising miniproteins against targets selected by Astellas. The biotechnology company may also receive additional payments when agreed development and commercial milestones are achieved, along with royalties on net sales of products that reach the market. The companies have not disclosed the upfront payment or the individual milestone amounts.

Chris Bahl, president, chief executive officer and founder of AI Proteins, said the collaboration would help advance the company’s approach to designing medicines from scratch. He highlighted Astellas’ experience in turning scientific research into treatments as an important factor in the partnership.

Morten Sogaard, senior vice president at Astellas Innovation Lab, said the agreement supports the company’s strategy of strengthening its early-stage research pipeline through partnerships and emerging technologies.

For Astellas, the collaboration provides access to an AI-based protein design platform that could expand its options for discovering new medicines. The partnership also reflects growing pharmaceutical industry interest in using artificial intelligence to improve early-stage research and identify potential drug candidates.

For AI Proteins, the agreement offers an opportunity to work with a global pharmaceutical company while advancing its technology towards potential treatments. The company is based in Boston and focuses on developing computationally designed miniproteins for therapeutic applications.

Although the collaboration represents a significant investment in AI-driven drug discovery, the targets and therapeutic areas remain undisclosed. Any medicines resulting from the partnership will need to undergo further development and regulatory review before they can become available to patients.