Efficient Sampling of Ligand Orientations and Conformations in Free Energy Calculations Using the λ-Dynamics Method

Abstract

The recently developed λ-dynamics free-energy based simulation method was used to study the binding of 10 five-member ring heterocycle derivatives to an artificial cavity created by mutagenesis inside cytochrome c peroxidase. Application of λ-dynamics to this system gives a reasonable estimate of the binding affinity of the ligands. This methodology also provides a means to explore the binding orientations and conformations of the ligands inside the binding pocket much better than does conventional MD. This is due to the scaling of forces inherent in the λ-dynamics method, which lowers the barriers separating different binding modes and conformations. Examination of the λ-dynamics trajectory of the ligands revealed alternative binding orientations and conformations not detected by crystallography. Furthermore, a λ-dynamics simulation starting from random initial orientations, in which some ligands take significantly different orientations as compared with those from the X-ray structure, successfully samples the X-ray crystallographic orientations in all ligands. Ligand sampling by conventional MD starting from same initial structures remains trapped in the local minima from which they start. Such efficient sampling of ligand orientations and conformations is expected to diminish the limitation that an initial ligand structure must be close to its true bound orientation in order to yield a reasonable estimate of the binding free energy.

Predicting Binding Free Energies: Frontiers and Benchmarks

“…Among host-guest systems, there is a particularly extensive experimental literature on cyclodextrins (46, 101), and they are tractable computationally (50, 132). As to artificial protein binding sites, the two variants of the CCP protein model binding site (34, 4, 5, 103, 95, 106) offer a modest increase in difficulty relative to the T4 lysozyme sites discussed above. And thrombin and the bromodomains appear to be promising examples of candidate drug targets for inclusion in a growing set of benchmark systems.…”

Perspective: Alchemical free energy calculations for drug discovery

“…Other applications have highlighted similar issues. 18,[109][110][111][112] Water plays a thermodynamically significant role as well-slow water motions into and out of binding sites can yield errors in computed relative free energies in excess of 10 kcal/mol. 87 Thus, available computational data, though limited, suggest that issues relating to uncertain, incorrect, or changing ligand binding modes can introduce errors up to 7 kcal/mol in relative binding free energy calculations when these effects are ignored.…”

Recent Advances in Free Energy Calculations with a Combination of Molecular Mechanics and Continuum Models

“…1 [4,[6][7][8][9], this review mainly focuses on the publications after the year of 2001. We also want to point out that there are many theoretical approaches for free energy calculations, which include free energy perturbation [17][18][19][20][21][22][23], linear interaction energy [26][27][28][29][30], hybrid-LIE/GB [31][32], hybrid-LIE/PB [33], -dynamics [34][35], the generalized Bjerrum approach [36] and ligand interaction scanning [37], doublecoupling method [38], potential mean force based approach [39], wormhole Monte Carlo method [40], statistical mechanical method based on molecular correlation functions [41] and so on. The comparisons of these methods to MM-PBSA and MM-GBSA are beyond the scope of this paper.…”