Pharma Giants Unite to Train AI for Next-Gen Drug Discovery

Post by : Bianca Suleiman

In a groundbreaking collaboration, Bristol Myers Squibb, Takeda Pharmaceuticals, and Astex Pharmaceuticals are pooling their proprietary data to train an advanced AI model, OpenFold3, designed to accelerate drug discovery. The move marks a significant step toward leveraging artificial intelligence for faster, more precise identification of protein-small molecule interactions.

The initiative operates on a federated data-sharing platform developed by Apheris, which allows companies to combine diverse datasets securely without exposing sensitive information. By aggregating thousands of experimentally determined protein–small molecule structures, the AI model aims to improve prediction accuracy for potential drug candidates.

OpenFold3 is part of the AI Structural Biology Network, an industry-led effort in collaboration with the AlQuraishi Lab at Columbia University. The consortium’s approach is designed to maximize efficiency and innovation, enabling breakthroughs that single organizations might struggle to achieve alone.

This collaboration reflects a larger trend in pharmaceutical research, where AI-driven solutions are increasingly becoming central to discovering new treatments and improving patient outcomes. With multiple global players joining forces, OpenFold3 could set new standards for predictive modeling in drug development.