Researchers at the Wadhwani School of Artificial Intelligence and Data Science in the Indian Institute of Technology-Madras (IIT-M) and The Ohio State University have managed to develop an artificial intelligence (AI) framework that can generate drug-like molecules that are easier to synthesise in a laboratory.

The framework is being called Policy-guided Unbiased Representations for Structure Constrained Molecular Generation or PURE. This framework promises to cut down early stage timelines of drug development, which is sometimes a decades-long process that costs millions of dollars. It could play a role in addressing drug resistance in cancer and infectious diseases, a press release said.

This, the researchers stated, has helped solve one of the biggest problems in AI-driven discovery where most molecules look promising on a computer but are nearly impossible to synthesise in reality. PURE grounds molecular generation in real synthesis pathways, automatically learns chemical similarity instead of relying on metrics that introduce bias, and suggests viable synthetic routes along with molecular structures. This would result in faster drug pipelines and back up solutions for failing treatments, the release added.

The findings were published in the Journal for Cheminformatics, an open-access journal that publishes original peer-reviewed research papers.

B. Ravindran, head, Wadhwani School of Data Science and AI (WSAI), IIT Madras, said PURE had moved researchers closer to AI systems that could reason through synthesis steps much like a chemist. Karthik Raman of WSAI said PURE adopted an approach of mapping chemical space without being biased towards a specific metric. Srinivasan Parthasarathy of Department of Computer Science and Engineering, The Ohio State University, said PURE offered early-stage discovery benefits for pharmaceutical research with the capability to identify alternative drug candidates in the face of resistance and hepatoxicity.

Published – November 03, 2025 11:42 pm IST