Improving molecular docking through eHiTS’ tunable scoring function

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.     

Machine learning optimization of cross docking accuracy

Performance of small molecule automated docking programs has conceptually been divided into docking -, scoring -, ranking - and screening power, which focuses on the crystal pose prediction, affinity prediction, ligand ranking and database screening capabilities of the docking program, respectively. Benchmarks show that different docking programs can excel in individual benchmarks which suggests that the scoring function employed by the programs can be optimized for a particular task. Here the scoring function of Smina is re-optimized towards enhancing the docking power using a supervised machine learning approach and a manually curated database of ligands and cross docking receptor pairs. The optimization method does not need associated binding data for the receptor-ligand examples used in the data set and works with small train sets. The re-optimization of the weights for the scoring function results in a similar docking performance with regard to docking power towards a cross docking test set. A ligand decoy based benchmark indicates a better discrimination between poses with high and low RMSD. The reported parameters for Smina are compatible with Autodock Vina and represent ready-to-use alternative parameters for researchers who aim at pose prediction rather than affinity prediction.     

Cross-docking study on InhA inhibitors: a combination of Autodock Vina and PM6-DH2 simulations to retrieve bio-active conformations

InhA, the NADH-dependent enoyl-acyl carrier protein reductase from Mycobacterium tuberculosis (Mtb) is the proposed main target of the first-line antituberculosis drug isoniazid (INH). INH activity is dependent on activation by the catalase peroxidase KatG, a Mtb enzyme whose mutations are linked to clinical resistance to INH. Other inhibitors of InhA that do not require any preliminary activation are known. The design of such direct potent inhibitors represents a promising approach to circumvent this resistance mechanism. An ensemble-docking process with four known InhA X-ray crystal structures and employing the Autodock Vina software was performed. Five InhA inhibitors whose bioactive conformations are known were sequentially docked in the substrate cavity of each protein. The efficiency of the docking was assessed and validated by comparing the calculated conformations to the crystallographic structures. For a same inhibitor, the docking results differed from one InhA conformation to another; however, docking poses that matched correctly or were very close to the expected bioactive conformations could be identified. The expected conformations were not systematically well ranked by the Autodock Vina scoring function. A post-docking optimization was carried out on all the docked conformations with the AMMP force field implemented on the VEGAZZ software, followed by a single point calculation of the interaction energy, using the MOPAC PM6-DH2 semi-empirical quantum chemistry method. The conformations were subsequently submitted to a PM6-DH2 optimization in partially flexible cavities. The resulting interaction energies combined with the multiple receptor conformations approach allowed us to retrieve the bioactive conformation of each ligand.     

Binding energy landscape analysis helps to discriminate true hits from high-scoring decoys in virtual screening

Although virtual screening through molecular docking has been widely applied in lead discovery, it is still challenging to distinguish true hits from high-scoring decoys because of the difficulty in accurately predicting protein-ligand binding affinities. Following the successful application of energy landscape analysis to both protein folding and biomolecular binding studies, we attempted to use protein-ligand binding energy landscape analysis to recognize true binders from high-scoring decoys. Two parameters describing the binding energy landscape were used for this purpose. The energy gap, defined as the difference between the binding energy of the native binding mode and the average binding energy of other binding modes in the "denatured binding phase", was used to describe the thermodynamic stability of binding, and the number of local binding wells in the landscapes was used to account for the kinetic accessibility. These parameters, together with the docking score, were combined using logistic regression to investigate their capability to discriminate true ligands from high-scoring decoys. Inhibitors and the noninhibitors of two enzyme systems, neuraminidase and cyclooxygenase-2, were used to test their discrimination capability. Using a five-fold cross-validation, the areas under the receiver operator characteristic curves (AUCs) from the best linear combinations of parameters reached 0.878 for neuraminidase and 0.776 for cyclooxygenase-2. To make a more independent test, inhibitors and high-scoring decoys in a directory of useful decoys (DUD), the largest and most comprehensive public data set for benchmarking virtual screen programs by far, were used as independent test sets to test the discrimination capability of these parameters. The AUCs of the best linear combinations of parameters for the independent test sets were 0.750 for neuraminidase and 0.855 for cyclooxygenase-2. Furthermore, combining these two parameters with the docking scoring function improved the enrichment ratio to 200-300% compared to that using the scoring function alone. This study suggests that incorporating information from binding energy landscape analysis can significantly increase the success rate of virtual screening.     

Evaluation of docking/scoring approaches: A comparative study based on MMP3 inhibitors

An increasing number of docking/scoring programs are available that use different sampling and scoring algorithms. A reliable scoring function is the crucial element of such approaches. Comparative studies are needed to evaluate their current capabilities. DOCK4 with force field and PMF scoring as well as FlexX were used to evaluate the predictive power of these docking/scoring approaches to identify the correct binding mode of 61 MMP-3 inhibitors in a crystal structure of stromelysin and also to rank them according to their different binding affinities. It was found that DOCK4/PMF scoring performs significantly better than FlexX and DOCK4/FF in both ranking ligands and predicting their binding modes. Most notably, DOCK4/PMF was the only scoring/docking approach that found a significant correlation between binding affinity and predicted score of the docked inhibitors. However, comparing only those cases where the correct binding mode was identified (scoring highest among sampled poses), FlexX showed the best `fine tuning' (lowest rmsd) in predicted binding modes. The results suggest that not so much the sampling procedure but rather the scoring function is the crucial element of a docking program.     

A D3R prospective evaluation of machine learning for protein-ligand scoring

We assess the performance of several machine learning-based scoring methods at protein-ligand pose prediction, virtual screening, and binding affinity prediction. The methods and the manner in which they were trained make them sufficiently diverse to evaluate the utility of various strategies for training set curation and binding pose generation, but they share a novel approach to classification in the context of protein-ligand scoring. Rather than explicitly using structural data such as affinity values or information extracted from crystal binding poses for training, we instead exploit the abundance of data available from high-throughput screening to approach the problem as one of discriminating binders from non-binders. We evaluate the performance of our various scoring methods in the 2015 D3R Grand Challenge and find that although the merits of some features of our approach remain inconclusive, our scoring methods performed comparably to a state-of-the-art scoring function that was fit to binding affinity data.     

Interaction with specific HSP90 residues as a scoring function: validation in the D3R Grand Challenge 2015

Here is reported the development of a novel scoring function that performs remarkably well at identifying the native binding pose of a subset of HSP90 inhibitors containing aminopyrimidine or resorcinol based scaffolds. This scoring function is called PocketScore, and consists of the interaction energy between a ligand and three residues in the binding pocket: Asp93, Thr184 and a water molecule. We integrated PocketScore into a molecular docking workflow, and used it to participate in the Drug Design Data Resource (D3R) Grand Challenge 2015 (GC2015). PocketScore was able to rank 180 molecules of the GC2015 according to their binding affinity with satisfactory performance. These results indicate that the specific residues considered by PocketScore are determinant to properly model the interaction between HSP90 and its subset of inhibitors containing aminopyrimidine or resorcinol based scaffolds. Moreover, the development of PocketScore aimed at improving docking power while neglecting the prediction of binding affinities, suggesting that accurate identification of native binding poses is a determinant factor for the performance of virtual screens.     

Energetic analysis of fragment docking and application to structure-based pharmacophore hypothesis generation

We have developed a method that uses energetic analysis of structure-based fragment docking to elucidate key features for molecular recognition. This hybrid ligand- and structure-based methodology uses an atomic breakdown of the energy terms from the Glide XP scoring function to locate key pharmacophoric features from the docked fragments. First, we show that Glide accurately docks fragments, producing a root mean squared deviation (RMSD) of <1.0 Å for the top scoring pose to the native crystal structure. We then describe fragment-specific docking settings developed to generate poses that explore every pocket of a binding site while maintaining the docking accuracy of the top scoring pose. Next, we describe how the energy terms from the Glide XP scoring function are mapped onto pharmacophore sites from the docked fragments in order to rank their importance for binding. Using this energetic analysis we show that the most energetically favorable pharmacophore sites are consistent with features from known tight binding compounds. Finally, we describe a method to use the energetically selected sites from fragment docking to develop a pharmacophore hypothesis that can be used in virtual database screening to retrieve diverse compounds. We find that this method produces viable hypotheses that are consistent with known active compounds. In addition to retrieving diverse compounds that are not biased by the co-crystallized ligand, the method is able to recover known active compounds from a database screen, with an average enrichment of 8.1 in the top 1% of the database.     

Identification of a minimal subset of receptor conformations for improved multiple conformation docking and two-step scoring

Docking and scoring are critical issues in virtual drug screening methods. Fast and reliable methods are required for the prediction of binding affinity especially when applied to a large library of compounds. The implementation of receptor flexibility and refinement of scoring functions for this purpose are extremely challenging in terms of computational speed. Here we propose a knowledge-based multiple-conformation docking method that efficiently accommodates receptor flexibility thus permitting reliable virtual screening of large compound libraries. Starting with a small number of active compounds, a preliminary docking operation is conducted on a large ensemble of receptor conformations to select the minimal subset of receptor conformations that provides a strong correlation between the experimental binding affinity (e.g., Ki, IC50) and the docking score. Only this subset is used for subsequent multiple-conformation docking of the entire data set of library (test) compounds. In conjunction with the multiple-conformation docking procedure, a two-step scoring scheme is employed by which the optimal scoring geometries obtained from the multiple-conformation docking are re-scored by a molecular mechanics energy function including desolvation terms. To demonstrate the feasibility of this approach, we applied this integrated approach to the estrogen receptor alpha (ERalpha) system for which published binding affinity data were available for a series of structurally diverse chemicals. The statistical correlation between docking scores and experimental values was significantly improved from those of single-conformation dockings. This approach led to substantial enrichment of the virtual screening conducted on mixtures of active and inactive ERalpha compounds.     

Customizing scoring functions for docking

Empirical scoring functions used in protein-ligand docking calculations are typically trained on a dataset of complexes with known affinities with the aim of generalizing across different docking applications. We report a novel method of scoring-function optimization that supports the use of additional information to constrain scoring function parameters, which can be used to focus a scoring function’s training towards a particular application, such as screening enrichment. The approach combines multiple instance learning, positive data in the form of ligands of protein binding sites of known and unknown affinity and binding geometry, and negative (decoy) data of ligands thought not to bind particular protein binding sites or known not to bind in particular geometries. Performance of the method for the Surflex-Dock scoring function is shown in cross-validation studies and in eight blind test cases. Tuned functions optimized with a sufficient amount of data exhibited either improved or undiminished screening performance relative to the original function across all eight complexes. Analysis of the changes to the scoring function suggest that modifications can be learned that are related to protein-specific features such as active-site mobility.     

Improving VEGFR-2 docking-based screening by pharmacophore postfiltering and similarity search postprocessing

Conventional docking-based virtual screening (VS) of chemical databases is based on the ranking of compounds according to the values retrieved by a scoring function (typically, the binding affinity estimation). However, using the most suitable scoring function for each kind of receptor pocket is not always an effective process to rank compounds, and sometimes neither to distinguish between correct binding modes from incorrect ones. To improve actives from decoys selection, here we propose a three-step VS protocol, which includes the conventional docking step, a pharmacophore postfilter step, and a similarity search postprocess. This VS protocol is retrospectively applied to VEGFR-2 (Kdr-kinase) inhibitors. The resulting docking poses calculated using the Alpha HB scoring function implemented in MOE are postfiltered according to defined pharmacophore interactions (structure based). The selected poses are again ranked according to their molecular similarity (MACCS fingerprint) to the cognate ligand. Results show that both the overall and early VS performance improve the application of this protocol.     

Exhaustive docking and solvated interaction energy scoring: lessons learned from the SAMPL4 challenge

We continued prospective assessments of the Wilma–solvated interaction energy (SIE) platform for pose prediction, binding affinity prediction, and virtual screening on the challenging SAMPL4 data sets including the HIV-integrase inhibitor and two host–guest systems. New features of the docking algorithm and scoring function are tested here prospectively for the first time. Wilma–SIE provides good correlations with actual binding affinities over a wide range of binding affinities that includes strong binders as in the case of SAMPL4 host–guest systems. Absolute binding affinities are also reproduced with appropriate training of the scoring function on available data sets or from comparative estimation of the change in target’s vibrational entropy. Even when binding modes are known, SIE predictions lack correlation with experimental affinities within dynamic ranges below 2 kcal/mol as in the case of HIV-integrase ligands, but they correctly signaled the narrowness of the dynamic range. Using a common protein structure for all ligands can reduce the noise, while incorporating a more sophisticated solvation treatment improves absolute predictions. The HIV-integrase virtual screening data set consists of promiscuous weak binders with relatively high flexibility and thus it falls outside of the applicability domain of the Wilma–SIE docking platform. Despite these difficulties, unbiased docking around three known binding sites of the enzyme resulted in over a third of ligands being docked within 2 Å from their actual poses and over half of the ligands docked in the correct site, leading to better-than-random virtual screening results.     

Can we trust docking results? Evaluation of seven commonly used programs on PDBbind database

Docking is one of the most commonly used techniques in drug design. It is used for both identifying correct poses of a ligand in the binding site of a protein as well as for the estimation of the strength of protein–ligand interaction. Because millions of compounds must be screened, before a suitable target for biological testing can be identified, all calculations should be done in a reasonable time frame. Thus, all programs currently in use exploit empirically based algorithms, avoiding systematic search of the conformational space. Similarly, the scoring is done using simple equations, which makes it possible to speed up the entire process. Therefore, docking results have to be verified by subsequent in vitro studies. The purpose of our work was to evaluate seven popular docking programs (Surflex, LigandFit, Glide, GOLD, FlexX, eHiTS, and AutoDock) on the extensive dataset composed of 1300 protein–ligands complexes from PDBbind 2007 database, where experimentally measured binding affinity values were also available. We compared independently the ability of proper posing [according to Root mean square deviation (or Root mean square distance) of predicted conformations versus the corresponding native one] and scoring (by calculating the correlation between docking score and ligand binding strength). To our knowledge, it is the first large‐scale docking evaluation that covers both aspects of docking programs, that is, predicting ligand conformation and calculating the strength of its binding. More than 1000 protein–ligand pairs cover a wide range of different protein families and inhibitor classes. Our results clearly showed that the ligand binding conformation could be identified in most cases by using the existing software, yet we still observed the lack of universal scoring function for all types of molecules and protein families. © 2010 Wiley Periodicals, Inc. J Comput Chem, 2011     

Comparison of two implementations of the incremental construction algorithm in flexible docking of thrombin inhibitors

A set of 32 known thrombin inhibitors representing different chemical classes has been used to evaluate the performance of two implementations of incremental construction algorithms for flexible molecular docking: DOCK 4.0 and FlexX 1.5. Both docking tools are able to dock 10–35% of our test set within 2 Å of their known, bound conformations using default sampling and scoring parameters. Although flexible docking with DOCK or FlexX is not able to reconstruct all native complexes, it does offer a significant improvement over rigid body docking of single, rule-based conformations, which is still often used for docking of large databases. Docking of sets of multiple conformers of each inhibitor, obtained with a novel protocol for diverse conformer generation and selection, yielded results comparable to those obtained by flexible docking. Chemical scoring, which is an empirically modified force field scoring method implemented in DOCK 4.0, outperforms both interaction energy scoring by DOCK and the Böhm scoring function used by FlexX in rigid and flexible docking of thrombin inhibitors. Our results indicate that for reliable docking of flexible ligands the selection of anchor fragments, conformational sampling and currently available scoring methods still require improvement.     

A novel scoring function for molecular docking

We present a novel scoring function for docking of small molecules to protein binding sites. The scoring function is based on a combination of two main approaches used in the field, the empirical and knowledge-based approaches. To calibrate the scoring function we used an iterative procedure in which a ligand's position and its score were determined self-consistently at each iteration. The scoring function demonstrated superiority in prediction of ligand positions in docking tests against the commonly used Dock, FlexX and Gold docking programs. It also demonstrated good accuracy of binding affinity prediction for the docked ligands.     

Knowledge-based interaction fingerprint scoring: a simple method for improving the effectiveness of fast scoring functions

A new method for the postprocessing of docking outputs has been developed, based on encoding putative 3D binding modes (docking solutions) as ligand-protein interactions into simple bit strings, a method analogous to the structural interaction fingerprint. Instead of employing traditional scoring functions, the method uses a series of new, knowledge-based scores derived from the similarity of the bit strings for each docking solution to that of a known reference binding mode. A GOLD docking study was carried out using the Bissantz estrogen receptor antagonist set along with the new scoring method. Superior recovery rates, with up to 2-fold enrichments, were observed when the new knowledge-based scoring was compared to the GOLD fitness score. In addition, top ranking sets of molecules (actives and potential actives or decoys) were structurally diverse with low molecular weights and structural complexities. Principal component analysis and clustering of the fingerprints permits the easy separation of active from inactive binding modes and the visualization of diverse binding modes.     

Scoring noncovalent protein-ligand interactions: A continuous differentiable function tuned to compute binding affinities

Summary Exploitation of protein structures for potential drug leads by molecular docking is critically dependent on methods for scoring putative protein-ligand interactions. An ideal function for scoring must exhibit predictive accuracy and high computational speed, and must be tolerant of variations in the relative protein-ligand molecular alignment and conformation. This paper describes the development of an empirically derived scoring function, based on the binding affinities of protein-ligand complexes coupled with their crystallographically determined structures. The function's primary terms involve hydrophobic and polar complementarity, with additional terms for entropic and solvation effects. The issue of alignment/conformation dependence was solved by constructing a continuous differentiable nonlinear function with the requirement that maxima in ligand conformation/alignment space corresponded closely to crystallographically determined structures. The expected error in the predicted affinity based on cross-validation was 1.0 log unit. The function is sufficiently fast and accurate to serve as the objective function of a molecular-docking search engine. The function is particularly well suited to the docking problem, since it has spatially narrow maxima that are broadly accessible via gradient descent.     

Exponential repulsion improves structural predictability of molecular docking

Molecular docking is a powerful tool for theoretical prediction of the preferred conformation and orientation of small molecules within protein active sites. The obtained poses can be used for estimation of binding energies, which indicate the inhibition effect of designed inhibitors, and therefore might be used for in silico drug design. However, the evaluation of ligand binding affinity critically depends on successful prediction of the native binding mode. Contemporary docking methods are often based on scoring functions derived from molecular mechanical potentials. In such potentials, nonbonded interactions are typically represented by electrostatic interactions between atom‐centered partial charges and standard 6–12 Lennard–Jones potential. Here, we present implementation and testing of a scoring function based on more physically justified exponential repulsion instead of the standard Lennard–Jones potential. We found that this scoring function significantly improved prediction of the native binding modes in proteins bearing narrow active sites such as serine proteases and kinases. © 2016 Wiley Periodicals, Inc.     

Novel, customizable scoring functions, parameterized using N-PLS, for structure-based drug discovery

The ability to accurately predict biological affinity on the basis of in silico docking to a protein target remains a challenging goal in the CADD arena. Typically, "standard" scoring functions have been employed that use the calculated docking result and a set of empirical parameters to calculate a predicted binding affinity. To improve on this, we are exploring novel strategies for rapidly developing and tuning "customized" scoring functions tailored to a specific need. In the present work, three such customized scoring functions were developed using a set of 129 high-resolution protein-ligand crystal structures with measured Ki values. The functions were parametrized using N-PLS (N-way partial least squares), a multivariate technique well-known in the 3D quantitative structure-activity relationship field. A modest correlation between observed and calculated pKi values using a standard scoring function (r2 = 0.5) could be improved to 0.8 when a customized scoring function was applied. To mimic a more realistic scenario, a second scoring function was developed, not based on crystal structures but exclusively on several binding poses generated with the Flo+ docking program. Finally, a validation study was conducted by generating a third scoring function with 99 randomly selected complexes from the 129 as a training set and predicting pKi values for a test set that comprised the remaining 30 complexes. Training and test set r2 values were 0.77 and 0.78, respectively. These results indicate that, even without direct structural information, predictive customized scoring functions can be developed using N-PLS, and this approach holds significant potential as a general procedure for predicting binding affinity on the basis of in silico docking.