New AI algorithm could aid drug discovery

Researchers at King’s College London and Imperial College London have developed an AI algorithm, DrugSynthMC, that can rapidly generate thousands of virtual drug-like molecules tailored to specific targets. This tool enhances early drug discovery by creating new chemical structures beyond existing virtual libraries, overcoming limitations that only allowed screening of pre-existing compounds. DrugSynthMC, which uses a Monte Carlo Tree Search algorithm, is highly efficient and doesn’t rely on training datasets. It generates molecules that are easy to synthesise, soluble, and non-toxic, and can be customised for specific biological targets, offering potential for advancing AI-driven drug discovery.

Could this breakthrough revolutionise the future of personalised medicine?

1 Like