Technological Trajectories and Future Roadmap of AI-Driven Drug Discovery: A Structural Topic Modeling Analysis of Global Patent Data

Artificial intelligence (AI) is radically transforming the pharmaceutical industry by converting protracted, trial-and-error drug discovery into rapid, precise, cost-effective, and data-driven approaches for molecular discovery, target identification, biomarker discovery, virtual screening, clinical trial optimization, and personalized therapeutics. Although research on AI-driven drug discovery has experienced substantial growth in recent years, the underlying innovation trajectories shaping this field remain insufficiently understood, as limited studies have mapped its technological structure and future evolutionary pathways. Moreover, the current literature lacks a future-oriented patent intelligence perspective on AI-driven drug discovery to uncover latent themes and emerging innovation hotspots. Addressing this gap, the present study leverages Structural Topic Modeling (STM) to identify and interpret the evolving technological landscape of AI-driven drug discovery from the patent dataset comprising 1078 patents from 2006 to 2025. The study identifies eight major themes, including deep learning for target identification and drug-target interaction modeling, AI-enabled in silico drug development and drug activity modeling, AI for clinical diagnostics and biomarker discovery, AI-assisted inhibitor discovery, AI for binding affinity prediction and molecular docking, precision medicine and AI-based treatment response prediction, intelligent drug screening, and generative AI for drug design. The analysis reveals the technological evolution from conventional computational drug discovery to generative AI-based design of high-potential compounds, target identification, virtual screening, precision therapeutics, and intelligent clinical trials. Further, the latent topics discovered using STM are classified into emerging, mature, and potentially stagnating themes to prepare a near-term, mid-term, and long-term technological roadmap of AI-driven drug discovery. The theoretical, practical, and strategic implications for academic researchers, policymakers, and innovation strategists are highlighted.