Convergence-Adaptive\nRoundtrip Method Enables Rapid\nand Accurate FEP Calculations

The free energy perturbation (FEP) method is a powerful\ntechnique\nfor accurate binding free energy calculations, which is crucial for\nidentifying potent ligands with a high affinity in drug discovery.\nHowever, the widespread application of FEP is limited by the high\ncomputational cost required to achieve equilibrium sampling and the\nchallenges in obtaining converged predictions. In this study, we present\nthe convergence-adaptive roundtrip (CAR) method, which is an enhanced\nadaptive sampling approach, to address the key challenges in FEP calculations,\nincluding the precision-efficiency tradeoff, sampling efficiency,\nand convergence assessment. By employing on-the-fly convergence analysis\nto automatically adjust simulation times, enabling efficient traversal\nof the important phase space through rapid propagation of conformations\nbetween different states and eliminating the need for multiple parallel\nsimulations, the CAR method increases convergence and minimizes computational\noverhead while maintaining calculation accuracy. The performance of\nthe CAR method was evaluated through relative binding free energy\n(RBFE) calculations on benchmarks comprising four diverse protein–ligand\nsystems. The results demonstrated a significant speedup of over 8-fold\ncompared to conventional FEP methods while maintaining high accuracy.\nThe overall <i>R</i><sup><i>2</i></sup> values\nof 0.65 and 0.56 were obtained using the combined-structure FEP approach\nand the single-step FEP approach, respectively, in conjunction with\nthe CAR method. In-depth case studies further highlighted the superior\nperformance of the CAR method in terms of convergence acceleration,\nimproved predicted correlations, and reduced computational costs.\nThe advancement of the CAR method makes it a highly effective approach,\nenhancing the applicability of FEP in drug discovery.