Scientists at the University of California, San Diego have developed an innovative method using artificial intelligence to simulate complex chemical processes, thus accelerating the discovery of new drugs.
Their tool, named POLYGON, is capable of quickly identifying drug candidates with multiple targets, which could reduce the side effects of current combined therapies.
The researchers used POLYGON to synthesize 32 new drug candidates aimed at treating cancer. This new tool is unique as it focuses on molecules that can act on multiple targets, an approach of great interest to scientists and physicians.
Algorithmics is now a growing field within pharmaceutical sciences, as explained by Trey Ideker, a professor at UC San Diego. Unlike commercial methods, this technology is made available as open source, allowing free use by everyone.
POLYGON was trained on a database of over a million bioactive molecules with detailed information about their chemical properties and known interactions. This technology enables the generation of chemical formulas for new drug candidates likely to inhibit specific proteins.
The researchers tested this technology by producing hundreds of drug candidates targeting various pairs of cancer-related proteins. Among them, 32 molecules showed promising interactions with the proteins MEK1 and mTOR, often targeted in combined therapies.
Drug discovery assisted by artificial intelligence is seen by researchers as a promising future path, potentially capable of transforming current practices.