UC San Diego's New AI Platform Accelerates Cancer Drug Development

UC San Diego scientists have made a breakthrough in cancer drug development with a groundbreaking AI platform called POLYGON. This innovative tool has the potential to revolutionize the drug discovery process, significantly cutting down on time and resources typically required.

Traditionally, identifying potential drugs involves an arduous process of thousands of experiments. However, with POLYGON, researchers can achieve comparable results in a fraction of the time. By simulating complex chemistry through machine learning, the platform has generated 32 novel drug candidates for cancer treatment.

What sets POLYGON apart is its ability to identify molecules targeting multiple proteins simultaneously, unlike conventional protocols that focus on single targets. This capability is crucial as multi-target drugs offer the promise of effective treatment with fewer side effects, akin to combination therapy.

Senior author Trey Ideker highlights the shift towards embracing AI in pharmaceuticals. Once viewed sceptically, AI is now integral to drug discovery, with startups incorporating it into their business plans to attract funding. The openness of UC San Diego's technology, in contrast to proprietary methods, democratizes access to AI-guided drug discovery.

POLYGON's training on a vast database of bioactive molecules enables it to generate original compounds tailored to desired chemical properties. In essence, it's akin to instructing the AI on how we want our future drug to interact with disease proteins, much like generating drawings based on desired characteristics.

The platform's efficacy was demonstrated by synthesizing 32 molecules targeting MEK1 and mTOR proteins, crucial in cancer therapy. These drugs exhibited significant activity against the proteins with minimal off-target effects, indicating promise for further development.

Ideker emphasizes the importance of human expertise in refining candidate drugs. While AI expedites certain steps, human chemists play a vital role in optimizing treatments.

Looking ahead, Ideker is optimistic about the potential of AI in drug discovery, foreseeing exciting developments in academia and the private sector in the coming decade.