PROTACs in targeted protein degradation: Advances in development and AI-enhanced drug discovery

Targeted protein degradation (TPD) eliminates disease-relevant proteins by engaging endogenous proteolytic machinery, most prominently the ubiquitin-proteasome system (UPS). Proteolysis-targeting chimeras (PROTACs) are heterobifunctional molecules composed of a protein of interest (POI) ligand, an E3 ligase recruiter, and a linker. By bringing the POI and E3 ligase into proximity, PROTACs promote POI ubiquitination and subsequent proteasomal degradation, and multiple candidates have progressed into clinical trials. This review summarizes the structural and mechanistic principles that govern PROTAC efficacy, selectivity, and degradation kinetics, and highlights key modality innovations and representative clinical progress with an emphasis on chemical structures, quantitative degradation metrics, and structure-activity relationships. We then examine key translational bottlenecks, including ternary-complex (TC) dependence, the hook effect, limited E3 ligase options, context-dependent selectivity, permeability, and beyond-Rule-of-Five (bRo5) properties, and discuss practical medicinal chemistry strategies to address these challenges. Finally, we describe how computational modeling and AI can be integrated across the design-make-test cycle, and summarize emerging data resources that enable more prospective, data-driven PROTAC discovery. AI-driven modular design framework for PROTAC development, integrating AI-based Data Processing, Ligand Discovery (POI ligands/warheads and E3 recruiters), POI Prioritization, Ternary Complex Modeling & Scoring, and AI-Driven Linker Optimization to generate PROTAC Assembly Blocks and assemble In Silico PROTAC Designs for downstream prioritization and experimental validation. • 1Summarizes mechanistic and structural principles underlying PROTAC-induced degradation. • 2Distills medicinal-chemistry strategies that tune ternary-complex cooperativity and degradation selectivity. • 3Provides a structured view of key translational limitations (hook effect, permeability, drug-likeness, E3 context, and bRo5). • 4Reviews computational and AI tools that mitigate major bottlenecks across the PROTAC workflow. • 5Discusses data resources and future directions for scalable, data-driven PROTAC discovery.