Target discovery and drug design in the era of artificial intelligence

Abstract The identification of therapeutically actionable targets and the design of effective small-molecule modulators remain central challenges in drug discovery, particularly for complex diseases driven by dynamic, interconnected molecular networks. Recent advances in artificial intelligence (AI) and machine learning are reshaping this landscape by enabling integrative analysis of large-scale biological data, systematic navigation of chemical space, and data-driven optimization of molecular properties. In this review, we outline how AI transforms small-molecule drug discovery by linking target discovery, tractability assessment, and chemical design within integrated, experiment-informed workflows. We highlight AI-enabled strategies for target discovery that integrate network biology, multimodal data fusion, and perturbation-aware modeling. These approaches identify context-specific, functionally actionable targets, including non-enzymatic proteins, protein-protein interactions, and other historically challenging classes. We then focus on advances most relevant to medicinal chemistry, including structure and complex prediction, docking and affinity modeling, ligand-based learning, generative molecular design, ADMET prediction, and synthetic feasibility assessment. Finally, we discuss current limitations related to data quality, generalizability, interpretability, and experimental translation. We emphasize the critical future role of integrated, context-aware AI workflows that connect target discovery with chemical design and experimental feedback. Together, these advances position AI as a robust framework for accelerating the discovery of therapeutically relevant targets and small-molecule modulators.