AI for Drug Discovery: From Algorithms to Medicines

Abstract

The growing integration of artificial intelligence with molecular dynamics simulations has introduced new possibilities for fostering and improving the discovery of inhibitors in drug development. This study aimed to systematically review how AI fosters MD. The researcher has focused on the methodological applications, reported benefits as compared with classical approaches, therapeutic contexts, and existing gaps. A structured review of peer-reviewed literature showed that AI has been integrated across various important domains. These integrations consistently improved accuracy, computational efficiency, and decision-making value. However, its applications span across kinases, viral proteins, and GPCRs. Although viral proteins and GPCRs have shown more mature applications, kinase-focused research appears comparatively limited. Despite these advances, notable challenges also persist, which include methodological opacity, lack of standardized benchmarks, and limited translational validation. The findings suggest that AI does not replace MD as it serves as a complementary approach that strengthens predictive and interpretive capacity. Thus, this review emphasizes the importance of standardized datasets, reproducible workflows, and experimental translation for maximizing the impact of AI-MD frameworks in drug discovery.