AI algorithm could aid drug discovery, find researchers
Researchers from King’s College London and Imperial College London have developed an AI algorithm called DrugSynthMC that has the potential to revolutionize drug discovery. This computer-based tool rapidly generates thousands of unique drug-like molecules, expanding the diversity of available compounds in libraries. Dr. Filippo Prischi, senior lecturer in molecular biochemistry at King’s College London, highlighted the algorithm’s ability to overcome limitations of existing drug collections by optimizing chemical diversity.
Virtual-library screening plays a crucial role in early drug discovery, using computational tools to identify compounds with high binding potential to specific drug targets. DrugSynthMC utilizes Monte Carlo Tree Search algorithm to predict outcomes and build structures of molecules with desired properties. The algorithm has shown success in generating compounds that are easy to synthesize, soluble, and non-toxic.
This groundbreaking approach, outlined in the Journal of Chemical Information and Modeling, is a significant advancement in AI-driven drug discovery. Dr. Olivier Pardo from Imperial College London praised the simplicity and efficiency of the algorithm, predicting its widespread use in identifying and optimizing compounds for therapeutic targets. The tool is publicly available for use by the research community, signaling a promising future for personalized drug discovery.
