Improving molecular docking through eHiTS’ tunable scoring function

Abstract

We present three complementary approaches for score-tuning that improve docking performance in pose prediction, virtual screening and binding affinity assessment. The methodology utilizes experimental data to customize the scoring function for the system of interest considering the specific docking scenario. The tuning approach, which has been implemented as an automated utility in eHiTS, is introduced as a solution to one of the conundrums of the molecular docking paradigm, namely, the lack of a universally well performing scoring function. The accuracy of scoring functions has been shown to be generally system-dependent, and particularly lacking for binding energy and bio-activity predictions. In the proposed approach, pose and energy predictions are enhanced by adjusting the relative weights of the eHiTS energy terms to improve score-RMSD or score-affinity correlations. In a virtual screening context ligand-based similarity is used to rescale the docking score such that better enrichment factors are achieved. We discuss the algorithmic details of the methods, and demonstrate the effects of score tuning on a variety of targets, including CDK2, BACE1 and neuraminidase, as well as on the popular benchmarks--the Directory of Useful Decoys and the PDBBind database.

Are predefined decoy sets of ligand poses able to quantify scoring function accuracy?

“…Given the lucky situation that structural and/or affinity data for the target at hand is available, assessing the performance of these scoring functions within the same docking package is obviously the most effective way for making an informed decision about which one to choose in the discovery process. Also the option of training a target-specific scoring function should be taken into account if enough training and test instances are available [60][61][62]. Although we in principle agree that a decoupling of sampling and scoring would be highly desirable, we believe that, for the reasons given in this study, the consideration of scoring function rankings derived from rescoring discrete decoy sets should be the last option in the decision making process.…”

Selective Small Molecule Compounds Increase BMP-2 Responsiveness by Inhibiting Smurf1-mediated Smad1/5 Degradation

“…Hence, in spite of spending much time on computing, the highly accurate computing result will surely reduce the demanded compounds amount in the second screening. We chose eHiTS to perform virtual screening because of its advantage in internal scoring system, which combines knowledge-based, statistics, the empirical approaches along with entropy loss estimation and grid based geometrical terms32. The major advantage of our in silico screening is: we enriched the initial compound pool by combining the product catalogs form four chemistry industrial companies, which ensured the chemical diversity that approx a million compounds still remained after fore-performed druggability selection.…”

Latest developments in molecular docking: 2010–2011 in review

“…The tunable scoring function of eHiTS [112] was tailored for a range of targets using experimental data. [113] Pose and energy estimates were enhanced by adjusting energy terms to improve score-RMSD and score-affinity correlations. GOLD and Glide were evaluated for pose prediction and VS performance against RNA targets, and it was shown that these "protein-based" docking programs can successfully dock into RNA and thus be useful in RNA-based drug design.…”