Big Pharma Unites to Train AI Model for Drug Discovery

In a rare show of collaboration, some of the world’s biggest pharmaceutical companies are pooling their strengths to train a cutting-edge artificial intelligence model designed to speed up drug discovery.

Bristol Myers Squibb, Takeda Pharmaceuticals and Astex Pharmaceuticals have joined a new consortium alongside AbbVie and Johnson & Johnson. Together, these firms will share data from thousands of experimentally determined protein–small molecule structures to enhance the training of OpenFold3, an advanced AI model built to predict how proteins interact with potential drug compounds.

Stay informed with the latest news. Follow DXB News Network on WhatsApp Channel

The initiative is powered by Germany-based Apheris, whose computing platform makes this large-scale collaboration possible. Instead of transferring sensitive data to a single database, Apheris enables federated data sharing—meaning every dataset stays securely within each company’s system while still contributing to the AI training.

By pooling this information, the consortium hopes to dramatically improve the accuracy of predictive models, helping researchers identify promising drug candidates more efficiently. The effort is part of the AI Structural Biology Network, an industry-driven project carried out in collaboration with Columbia University’s AlQuraishi Lab.

What makes this move notable is not just the technology but the willingness of major pharmaceutical rivals to work together. In an era where AI is increasingly central to healthcare innovation, this collaboration marks a step toward faster, more precise drug development that could benefit patients worldwide.