Evaluating the enthalpic contribution to ligand binding using QM calculations: effect of methodology on geometries and interaction energies

Abstract

As a result of research on ligand efficiency in the pharmaceutical industry, there is greater focus on optimizing the strength of polar interactions within receptors, so that the contribution of overall size and lipophilicity to binding can be decreased. A number of quantum mechanical (QM) methods involving simple probes are available to assess the H-bonding potential of different heterocycles or functional groups. However, in most receptors, multiple features are present, and these have distinct directionality, meaning very minimalist models may not be so ideal to describe the interactions. We describe how the use of gas phase QM models of kinase protein-ligand complex, which can more closely mimic the polar features of the active site region, can prove useful in assessing alterations to a core template, or different substituents. We investigate some practical issues surrounding the use of QM cluster models in structure based design (SBD). These include the choice of the method; semi-empirical, density functional theory or ab-initio, the choice of the basis set, whether to include implicit or explicit solvation, whether BSSE should be included, etc. We find a combination of the M06-2X method and the 6-31G* basis set is sufficiently rapid, and accurate, for the computation of structural and energetic parameters for this system.

Conformations and Binding Properties of Thiametoxam and Clothianidin Neonicotinoid Insecticides to Nicotinic Acetylcholine Receptors: The Contribution of σ‐Hole Interactions

“…DFT functionals, able to handle dispersion and stacking interactions, and to simulate properly H‐bonds, have been proven to achieve good accuracy and efficiency in large scale systems. The meta hybrid functional M06‐2X, designed by Zhao and Truhlar, has been successfully used in the description of non‐covalent interactions in biological systems . However, even if DFT methodologies require less computational resources than wave function based approaches, the full DFT description of models such as the one shown in Figure stays out of reach.…”

Binding of Sulfoxaflor to Aplysia californica-AChBP: Computational Insights from Multiscale Approaches

“…We have selected the M06-2X , functional associated with the 6-311G­(d) basis set, to perform our conformational investigations. Indeed, M06-2X has been proven to surpass other functionals for energy prediction and nonbonded interactions in the gas phase and in biological systems. − A full geometry optimization of the four SFX stereoisomers has first been carried out, and two stereoisomers ( SS and SR ) were considered for a systematic conformational analysis. In addition to gas-phase optimizations and to probe the effect of the surroundings, the located minima have also been fully optimized under the influence of solvent effects (dichloromethane and water), accounted for with the Solvation Model Density (SMD) .…”

Benchmark of electronic structure methods for protein–ligand interactions based on high-level reference data

“…4,5 Many commonly used \scoring functions" perform reasonably well for this task, but no single approach seems to be generally applicable when high accuracy is needed. 6,7 Especially the last¯ve years have seen a number of studies on how to improve upon these methods with computationally more demanding quantum chemistry based approaches, [8][9][10][11][12][13][14][15][16][17][18][19][20][21][22] but many questions are still open in this¯eld. We have recently published a large-scale study using systematically generated model systems from the PDBbind database 23,24 and comparing molecular mechanics (MM), semiempirical quantum mechanical (SQM) and density functional theory (DFT) methods.…”