Robust Scoring Functions for Protein–Ligand Interactions with Quantum Chemical Charge Models

Abstract

Ordinary least square (OLS) regression has been used widely for constructing the free scoring functions for protein-ligand interaction. However, OLS is very sensitive to the existence of outliers, and models constructed using it are easily affected by the outliers or even the choice of the dataset. On the other hand, determination of atomic charges is regarded as of central importance, because the electrostatic interaction is known to be a key contributing factor for biomolecular association. In the development of the AutoDock4 scoring function, only OLS was conducted, and the simple Gasteiger method was adopted. It is therefore of considerable interest to see whether more rigorous charge models could improve the statistical performance of the AutoDock4 scoring function. In this study, we have employed two well-established quantum chemical approaches, namely the restrained electrostatic potential (RESP) and the Austin-Model 1-Bond Charge Correction (AM1-BCC) methods, to obtain atomic partial charges, and we have compared how different charge models affect the performance of AutoDock4 scoring functions. In combination with robust regression analysis and outlier exclusion, our new protein-ligand free energy regression model with AM1-BCC charges for ligands and Amber99SB charges for proteins achieve lowest root-mean squared error of 1.637 kcal/mol for the training set of 147 complexes and 2.176 kcal/mol for the external test set of 1427 complexes. The assessment for binding pose prediction with the 100 external decoy sets indicates very high success rate of 87% with the criteria of predicted RMSD less than 2 Å. The success rates and statistical performance of our robust scoring functions are only weakly class-dependent (hydrophobic, hydrophilic, or mixed).

Disorder-to-helix conformational conversion of the human immunomodulatory peptide LL-37 induced by antiinflammatory drugs, food dyes and some metabolites

“…energies stem from the consideration of various physicochemical interactions found between the ligand and the helical peptide. It also needs to be stressed, that the semi-empirical Autodock4 force field was calibrated on ligand-protein complexes [38] instead of the much more flexible peptides, whereas conformational freedom, protonation state and charge distribution of the ligands can potentially affect the docking results [52,54]. Furthermore, the calculated binding free energies and dissociation constants are the averages of different docking poses as we had no structural information about the preferred binding sites.…”

Latest developments in molecular docking: 2010–2011 in review

“…Olson and co-workers investigated whether rigorous charge models (restrained electrostatic potential and Austin model 1 bond charge correction) could improve the statistical performance of AutoDock4. [107] Robust regression analysis with outlier exclusion Austin model 1 bond charge correction for ligands and AMBER99SB for proteins achieved a low error of prediction for affinity estimation and, more impressively, a very high success rate in pose prediction.…”

Review structure‐ and dynamics‐based computational design of anticancer drugs

“…Accuracy of prediction in pose and binding affinity remains the central issue in molecular docking, which was reflected in the relentless efforts in the makings of new scoring functions, as seen in some recent reviews on this subject . To interrogate the accuracy of computational software and methods, there are also contests for discriminating binders from non‐binders and for predicting binders’ binding poses and binding affinities on specific biomolecular targets, e.g., the HIV integrase in the SAMPL4 challenge .…”

“…A relatively economic solution to account for conformational flexibility of biomolecules with the rather efficient docking calculations is through the “relaxed complex” scheme, in which molecular dynamics simulations of the biomolecules are conducted to sample the conformations in the most relevant configuration subspace . It should be noted that the scoring functions to be incorporated in this scheme should be reasonably sensitive to conformational changes . In some recent applications it was shown that the predictive power of virtual screen against chemical library to find drug candidates can be improved with the relaxed complex scheme .…”