Selective Inhibitor Design for Kinase Homologs Using Multiobjective Monte Carlo Tree Search

Designing highly selective molecules for a drug target protein against off-targets, including pharmacokinetics and toxicity-related proteins, is a challenging task in drug discovery and can be regarded as a multiobjective problem. Recent breakthroughs in artificial intelligence have accelerated the development of molecular structure generation methods, and various researchers have applied them to computational drug designs and successfully proposed promising drug candidates. However, designing efficient selective inhibitors with releasing activities against various homologs of a target protein remains a difficult issue. In this study, we developed a de novo structure generator based on reinforcement learning that is capable of simultaneously optimizing multiobjective problems. Our structure generator successfully proposed selective inhibitors for 9 tyrosine-kinases while optimizing 9 other objectives related to pharmacokinetics and drug-likeness properties. These results show that our structure generator and optimization strategy for selective inhibitors will contribute to the further development of practical structure generators for drug designs.