Binding Free Energy simulations are widely used in early stage drug discovery to predict the binding affinities between lead candidates and their biomolecule targets. In this presentation, we introduce the application of Alchemical Enhanced Sampling (ACES) addressing the Relative Binding Free Energy (RBFE) and Core-hopping Binding Free Energy (CBFE) simulations, combined with a versatile analysis suit FE-Toolkit, specifically targeting the macrophage Migration Inhibitory Factor (MIF) system.

RBFE is traditionally applied to structurally similar ligands, efficiently quantifying the relative free energy differences within a congeneric molecule series by leveraging significant structural overlap. In contrast, CBFE is well-suited for enabling the estimation of binding free energies when no obvious chemical similarities exist between ligands. By combining RBFE and CBFE, we construct a perturbation network where RBFE provides dense connectivity within chemical series, while CBFE bridges distinct scaffolds. This hybrid approach enhances flexibility and applicability in drug discovery pipelines, where diverse ligand chemotypes are often explored simultaneously.

We integrate the optimized alchemical pathway and Alchemical Enhanced Sampling (ACES) method in all simulations to ensure sufficient sampling and accurate free energy estimates with improved phase space overlap and convergence. For CBFE, we apply λ-dependent Boresch-style restraints to explicitly control ligand positional and orientational degrees of freedom within the binding site. This is critical for challenging targets like MIF, which features a shallow and solvent-exposed binding pocket prone to sampling artifacts.

We demonstrate this framework on a series of MIF inhibitors spanning multiple scaffolds. The simulations capture key protein-ligand interactions and yield a Pearson correlation coefficient of 0.74 with experimental binding affinities. In addition, the ranking of ligands shows good agreement with experiment (R = 0.76), highlighting the method’s potential to prioritize compounds effectively in scaffold-level drug discovery efforts.

This work highlights the synergy of RBFE and CBFE for resolving complex thermodynamic networks in drug discovery, offering a scalable strategy to explore ligand series spanning multiple chemical scaffolds. Our findings provide practical guidance for applying enhanced alchemical sampling methods in systems like MIF and beyond.

Darrin York

Rutgers University

Appearances