Comparison of Receptor-Ligand Restraint Schemes for Alchemical Absolute Binding Free Energy Calculations

Abstract

Alchemical absolute binding free energy calculations are of increasing interest in drug discovery. These calculations require restraints between the receptor and ligand to restrict their relative positions and, optionally, orientations. Boresch restraints are commonly used, but they must be carefully selected in order to sufficiently restrain the ligand and to avoid inherent instabilities. Applying multiple distance restraints between anchor points in the receptor and ligand provides an alternative framework without inherent instabilities which may provide convergence benefits by more strongly restricting the relative movements of the receptor and ligand. However, there is no simple method to calculate the free energy of releasing these restraints due to the coupling of the internal and external degrees of freedom of the receptor and ligand. Here, a method to rigorously calculate free energies of binding with multiple distance restraints by imposing intramolecular restraints on the anchor points is proposed. Ab- solute binding free energies for the human macrophage migration inhibitory factor - MIF180 system obtained using a variety of Boresch restraints and rigorous and non- rigorous implementations of multiple distance restraints are compared. It is shown that several multiple distance restraints schemes produce estimates in good agreement with Boresch restraints. In contrast, calculations without orientational restraints produce erroneously favourable free energies of binding by up to approximately 4 kcal mol−1. These approaches offer new options for the deployment of alchemical absolute binding free energy calculations.

A nonequilibrium alchemical method for drug-receptor absolute binding free energy calculations: the role of restraints

“…, 51 respectively, where they observed that in case of solvent exposed binding pockets or side chain changes between the APO and HOLO structures longer alchemical transformations lead to better free energy estimates. In our case, the correct sampling of the slow degrees of freedom linked to the coupling/decoupling is in part assisted by the HREM sampling of the initial bound state, that helps with 'stiff' dihedrals, and in part by the long nonequilibrium alchemical transformations that tame the dissipation connected to the slow entrance of water solvent molecules in the unfettered cavity of the ligand at low coupling.…”

“…To see if restraining more the ligand during the annihilation gathers better results we also performed the calculations with the Loose-Tight restraint with two "frozen" atoms (see SI Figures 4, 5), the results obtained are similar to the one atom case but worse, this is in contradiction with what observed in ref. 12 where restraining the orientational degrees of freedom gathered better results. So we performed bidirectional calculations with one and two "frozen" atoms (see CI Figures 6,7,8,9), in general they gave worse results for twice the computational cost, but in this case the transformations with two "frozen" atoms, and therefore a restrained rotational degree of freedom, performed better than those with one "frozen" atom in agreement with ref.…”

“…Sluggish convergence with Boresch restraints, and even dependency on the arbitrarily selected anchor points to preserve the crystal pose, was recently observed by Michel and co-workers in their systematic analysis of ligand-receptor restraint schemes for FEP-based absolute binding free energies calculations of MIF180 in complex with human macrophage migration inhibitory factor. 51…”

“…Collective Variable (CV) based enhanced sampling algorithms. [4][5][6][7][8][9][10][11][12][13] Ease of use, speed, automation and reliability of the algorithm are particularly important in the context of drug discovery pipelines. In this respect, non-equilibrium alchemical techniques have been shown to strike a good balance [14][15][16][17] and be a valid alternative to traditional equilibrium alchemical techniques (FEP), 18 while being computationally cheaper than CV-based enhanced sampling algorithms (such as Metadynamics 19 ).…”

“…23 Albeit some progress has been made towards the automatic selection of restraints, the automatic set-up of the calculations is still problematic. 12,21,24 Recently we introduced the virtual double system single box vDSSB 25,26 method, which can be seen as a statistically boosted version of the monodirectional non-equilibrium work fast switching double annihilation method NEW-FSDAM. [27][28][29][30][31] This non-equilibrium technique differs from the more commonly used equilibrium and non-equilibrium alchemical techniques because it uses an Hamiltonian replica exchange with torsional tempering (HREM) [32][33][34] in order to get the Boltzmann distribution of the conformational landscape of the system, and also because the alchemical transformations are performed only in one direction.…”