Scientists are training AI to predict drug interactions

Editor's Note

An international team of scientists designed an artificial intelligence (AI) platform to provide insights into how a particular drug will interact in an individual prior to administration, Science Direct October 31 reports. The findings were published in the journal Cell Reports. 

Among the findings:

• One-third of drugs work by binding to cell-surface receptors called G protein-coupled receptors, or GPCRs, which have a “docking station” that allows drugs to penetrate the cell. 
• To create their platform, the scientists had to develop sophisticated data on how these GPCRs operate.
• The scientists used molecular tracking technology to better understand the actions of more than 100 leading cellular drug targets, as well as their primary genetic variations, and used that data to train an AI platform. 
• The algorithm predicted how cell surface receptors would respond to drug-like molecules with more than 80% accuracy 

The scientists say their ultimate goal is to pave the way for custom tailoring of prescriptions and precision medicine.