Bristol Myers, Takeda Join Forces on AI-Powered Drug Discovery Consortium

Bristol Myers Squibb, Takeda Pharmaceuticals, and Astex Pharmaceuticals announced a strategic collaboration to share proprietary molecular data for training an artificial intelligence model aimed at accelerating drug discovery and development.

The partnership joins a wider consortium that already includes major pharmaceutical players such as AbbVie and Johnson & Johnson, according to an announcement from Apheris, the German life sciences computing company facilitating the project.

FEDERATED AI FOR DRUG DEVELOPMENT

The companies will contribute data from several thousand experimentally determined protein–small molecule structures to train OpenFold3, an advanced AI model designed to predict protein-ligand interactions — a critical component in developing new medicines.

Unlike traditional data-sharing methods, this collaboration uses Apheris’ federated data platform, which allows multiple organizations to collaborate securely without exposing or transferring sensitive data. Each company’s dataset remains stored in its original location while being used to improve the AI model collectively.

“The federated approach allows us to advance predictive models for small molecule discovery in ways no single organization could achieve alone,” said Payal Sheth, Vice President of Discovery Biotherapeutics and Lead Discovery & Optimization at Bristol Myers Squibb.

A COLLABORATIVE AI STRUCTURAL BIOLOGY INITIATIVE

OpenFold3 serves as the flagship project of the AI Structural Biology Network, an industry-led consortium developed in collaboration with the AlQuraishi Lab at Columbia University. The initiative aims to push the boundaries of AI-driven molecular modeling and create a shared foundation for faster, more efficient therapeutic development.

Hans Bitter, Head of Computational Sciences at Takeda, said the collaboration fits seamlessly into the company’s broader digital strategy.
“This consortium really ties into our larger corporate goal of embedding AI throughout everything we do,” Bitter noted. “It’s a powerful example of how pharma companies can come together to achieve more for patients than we could individually.”

By combining their molecular and structural data through AI, the consortium partners hope to improve drug target identification, binding prediction accuracy, and development timelines, setting a new standard for AI-based biopharma collaboration.