A Korean research team has developed a technology that uses artificial intelligence (AI) to predict how cancer cells will respond to a drug they encounter for the first time.

Gwangju Institute of Science and Technology (GIST) announced on the 19th that a research team led by Professor Hyunju Lee of the Department of AI has developed an AI model called “InFo-DRP” that predicts cancer cell responses to new drugs.

Existing AI models have been effective at predicting responses to drugs they had been trained on, but faced limitations with reduced accuracy for previously unseen drugs. To address this, the research team changed the way the AI learns. The model is trained on the chemical structures of existing drugs, genetic information of cancer cells, and actual experimental results, and on that basis is designed to quickly identify key features even in drugs it has not seen before. For this purpose, the team analyzed drug structures and properties using multiple methods and applied an “invariant learning” technique to prevent over-reliance on the characteristics of any particular drug. The resulting AI model, InFo-DRP, predicts responses to new drugs by integrating drug characteristics with the genetic information of cancer cells.

Professor Lee said, “It is expected to be used in the early stages of new drug development to predict the responses of candidate substances in advance and to set priorities for experiments.” The research findings were published on August 30 in IEEE JBHI, an international journal in the field of biomedical and health informatics.

Popular News