MMPBSA

Molecular Mechanics Poisson-Boltzmann Surface Area (MM-PBSA) is a computational approach used to estimate the free energy changes associated with molecular binding events from molecular dynamics trajectories. It combines molecular mechanics energies with continuum solvent models to approximate the solvation contribution to binding free energy.

In typical MM-PBSA workflows, snapshots are taken from MD simulations of the complex and its separate components, such as the receptor and the ligand, in the same solvent environment. For each snapshot, the binding free energy is estimated by summing molecular mechanics energies (usually including electrostatics and van der Waals terms, and sometimes a reduced treatment of bonded terms) and solvation energies. The polar solvation energy is computed with either the Poisson-Boltzmann (PB) or Generalized Born (GB) model, while the nonpolar solvation energy is commonly modeled as a term proportional to the solvent-accessible surface area. The binding free energy is typically expressed as ΔG_bind ≈ ⟨E_complex⟩ − ⟨E_protein⟩ − ⟨E_ligand⟩ + ⟨G_solv,complex⟩ − ⟨G_solv,protein⟩ − ⟨G_solv,ligand⟩, with an entropy term, TΔS, often neglected or approximated due to computational cost.

MM-PBSA and MM-GBSA are common variants, differing mainly in how the polar solvation energy is calculated (PB vs GB). The method is relatively fast and can be applied to large systems, using MD sampling while avoiding the full cost of alchemical free-energy calculations. Limitations include dependence on implicit solvent models, sensitivity to parameter choices, and approximate treatment of conformational entropy, which can affect accuracy for systems with significant structural rearrangements or significant solvent effects. Applications include ranking ligands and estimating binding free energies for protein–protein and protein–ligand interactions.