Premium
New parameterization approaches of the LIE method to improve free energy calculations of PlmII‐Inhibitors complexes
Author(s) -
Valiente Pedro A.,
Gil L. Alejandro,
Batista Paulo R.,
Caffarena Ernesto R.,
Pons Tirso,
Pascutti Pedro G
Publication year - 2010
Publication title -
journal of computational chemistry
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.907
H-Index - 188
eISSN - 1096-987X
pISSN - 0192-8651
DOI - 10.1002/jcc.21566
Subject(s) - free energy perturbation , chemistry , perturbation theory (quantum mechanics) , perturbation (astronomy) , gauge (firearms) , thermodynamics , energy (signal processing) , computational chemistry , binding energy , term (time) , thermodynamic integration , approximation error , statistical physics , mathematics , physics , materials science , atomic physics , quantum mechanics , molecular dynamics , metallurgy
The standard parameterization of the Linear Interaction Energy (LIE) method has been applied with quite good results to reproduce the experimental absolute binding free energies for several protein–ligand systems. However, we found that this parameterization failed to reproduce the experimental binding free energy of Plasmepsin II (PlmII) in complexes with inhibitors belonging to four dissimilar scaffolds. To overcome this fact, we developed three approaches of LIE, which combine systematic approaches to predict the inhibitor‐specific values of α, β, and γ parameters, to gauge their ability to calculate the absolute binding free energies for these PlmII‐Inhibitor complexes. Specifically: (i) we modified the linear relationship between the weighted nonpolar desolvation ratio (WNDR) and the α parameter, by introducing two models of the β parameter determined by the free energy perturbation (FEP) method in the absence of the constant term γ, and (ii) we developed a new parameterization model to investigate the linear correlation between WNDR and the correction term γ. Using these parameterizations, we were able to reproduce the experimental binding free energy from these systems with mean absolute errors lower than 1.5 kcal/mol. © 2010 Wiley Periodicals, Inc. J Comput Chem 2010