From structure prediction to ligand–target recognition: expanding the reach of AlphaFold in natural product research

Abstract

Natural products constitute a structurally diverse reservoir of bioactive molecules and remain an important source of lead compounds for drug discovery. However, the limited availability of structural information on pharmacological targets and biomolecular interactions continues to hinder mechanism elucidation and structure-guided optimization, particularly for complex natural products such as bioactive peptides, proteinaceous natural products, and natural molecular glues. Recent advances in AlphaFold have transformed structural biology by enabling high-confidence prediction of protein structures and biomolecular complexes, providing unprecedented opportunities for structure-guided natural product research. Beyond conventional structure-based virtual screening, emerging co-folding strategies have demonstrated considerable potential for improving early enrichment in natural product screening and for elucidating the binding mechanisms of complex natural products with their targets. This review summarizes recent advances in AlphaFold-assisted natural product research, including target structure prediction, hit and lead discovery, co-folding-enabled interaction analysis, mechanistic investigation of peptide- and protein-derived natural products, and the emerging potential of AlphaFold for natural molecular glue discovery. We further discuss current limitations and future directions in structure prediction and interaction modeling. Overall, AlphaFold is reshaping natural product research from lead discovery to mechanistic characterization, with particular promise for advancing macromolecular natural products and molecular glue-mediated therapeutics. AlphaFold (AF) modeling is integrated into natural product (NP) drug discovery. AF-based target structural prediction promotes NP-derived hit-to-lead optimization. Advantages of AF3-based co-folding in early enrichment of screening are outlined. AF-driven macro-NP structural prediction and interaction with targets are reviewed. Challenges and future developments of AF in NP drug discovery are discussed. AlphaFold (AF) modeling is integrated into natural product (NP) drug discovery. AF-based target structural prediction promotes NP-derived hit-to-lead optimization. Advantages of AF3-based co-folding in early enrichment of screening are outlined. AF-driven macro-NP structural prediction and interaction with targets are reviewed. Challenges and future developments of AF in NP drug discovery are discussed.