FragScan: A Quantitative Fragment Scanning Strategy for Rational Drug Discovery

The calculation of binding free energy is a critical and challenging step in drug discovery and molecular design, as traditional methods often suffer from a trade-off between computational efficiency and prediction accuracy, and struggle to quantitatively analysis the contribution of individual ligand fragments to binding affinity. To address these limitations, this study proposes a novel fragment scanning approach-FragScan, which fragments ligand molecules at rotatable bonds, a strategy that aligns with the conformational flexibility of ligands and enables targeted analysis of fragment-receptor interactions. By decomposing ligands into structurally independent fragments, this method effectively reduces computational complexity while preserving high accuracy in energy calculations. Notably, it can accurately quantify the binding contribution of each ligand fragment, overcoming the drawback of conventional methods that fail to pinpoint fragment-specific effects. Our results demonstrate that FragScan provides a quantitatively reliable framework for predicting ligand fragment-receptor interactions, with validated performance in balancing efficiency and precision. This framework holds significant potential for advancing rational drug design, particularly in facilitating scaffold hopping and pharmacophore replacement-two core strategies for optimizing lead compounds and expanding chemical space. Collectively, FragScan offers a valuable tool for decoding structure-activity relationships at the fragment level, and is expected to drive progress in the development of novel and potent therapeutic agents.