Virtual Screening Performance Research Articles

Integrating Ligand-Based and Protein-Centric Virtual Screening of Kinase Inhibitors Using Ensembles of Multiple Protein Kinase Genes and Conformations

The rapidly growing wealth of structural and functional information about kinase genes and kinase inhibitors that is fueled by a significant therapeutic role of this protein family provides a significant impetus for development of targeted computational screening approaches. In this work, we explore an ensemble-based, protein-centric approach that allows for simultaneous virtual ligand screening against multiple kinase genes and multiple kinase receptor conformations. We systematically analyze and compare the results of ligand-based and protein-centric screening approaches using both single-receptor and ensemble-based docking protocols. A panel of protein kinase targets that includes ABL, EGFR, P38, CDK2, TK, and VEGFR2 kinases is used in this comparative analysis. By applying various performance metrics we have shown that ligand-centric shape matching can provide an effective enrichment of active compounds outperforming single-receptor docking screening. However, ligand-based approaches can be highly sensitive to the choice of inhibitor queries. Employment of multiple inhibitor queries combined with parallel selection ranking criteria can improve the performance and efficiency of ligand-based virtual screening. We also demonstrated that replica-exchange Monte Carlo docking with kinome-based ensembles of multiple crystal structures can provide a superior early enrichment on the kinase targets. The central finding of this study is that incorporation of the template-based structural information about kinase inhibitors and protein kinase structures in diverse functional states can significantly enhance the overall performance and robustness of both ligand and protein-centric screening strategies. The results of this study may be useful in virtual screening of kinase inhibitors potentially offering a beneficial spectrum of therapeutic activities across multiple disease states.

ALiBERO: evolving a team of complementary pocket conformations rather than a single leader.

Docking and virtual screening (VS) reach maximum potential when the receptor displays the structural changes needed for accurate ligand binding. Unfortunately, these conformational changes are often poorly represented in experimental structures or homology models, debilitating their docking performance. Recently, we have shown that receptors optimized with our LiBERO method (Ligand-guided Backbone Ensemble Receptor Optimization) were able to better discriminate active ligands from inactives in flexible-ligand VS docking experiments. The LiBERO method relies on the use of ligand information for selecting the best performing individual pockets from ensembles derived from normal-mode analysis or Monte Carlo. Here we present ALiBERO, a new computational tool that has expanded the pocket selection from single to multiple, allowing for automatic iteration of the sampling-selection procedure. The selection of pockets is performed by a dual method that uses exhaustive combinatorial search plus individual addition of pockets, selecting only those that maximize the discrimination of known actives compounds from decoys. The resulting optimized pockets showed increased VS performance when later used in much larger unrelated test sets consisting of biologically active and inactive ligands. In this paper we will describe the design and implementation of the algorithm, using as a reference the human estrogen receptor alpha.

Structure prediction and molecular dynamics simulations of a G-protein coupled receptor: human CCR2 receptor

CC chemokine receptor type-2 (CCR2) is a member of G-protein coupled receptors superfamily, expressed on the cell surface of monocytes and macrophages. It binds to the monocyte chemoattractant protein-1, a CC chemokine, produced at the sites of inflammation and infection. A homology model of human CCR2 receptor based on the recently available C-X-C chemokine recepor-4 crystal structure has been reported. Ligand information was used as an essential element in the homology modeling process. Six known CCR2 antagonists were docked into the model using simple and induced fit docking procedure. Docked complexes were then subjected to visual inspection to check their suitability to explain the experimental data obtained from site directed mutagenesis and structure-activity relationship studies. The homology model was refined, validated, and assessed for its performance in docking-based virtual screening on a set of CCR2 antagonists and decoys. The docked complexes of CCR2 with the known antagonists, TAK779, a dual CCR2/CCR5 antagonist, and Teijin-comp1, a CCR2 specific antagonist were subjected to molecular dynamics (MD) simulations, which further validated the binding modes of these antagonists. B-factor analysis of 20 ns MD simulations demonstrated that Cys190 is helpful in providing structural rigidity to the extracellular loop (EL2). Residues important for CCR2 antagonism were recognized using free energy decomposition studies. The acidic residue Glu291 from TM7, a conserved residue in chemokine receptors, is favorable for the binding of Teijin-comp1 with CCR2 by ΔG of −11.4 kcal/mol. Its contribution arises more from the side chains than the backbone atoms. In addition, Tyr193 from EL2 contributes −0.9 kcal/mol towards the binding of the CCR2 specific antagonist with the receptor. Here, the homology modeling and subsequent molecular modeling studies proved successful in probing the structure of human CCR2 chemokine receptor for the structure-based virtual screening and predicting the binding modes of CCR2 antagonists.

Impact of the CXCR4 structure on docking-based virtual screening of HIV entry inhibitors

Herein we analyze in depth the receptor-based virtual screening (VS) performance of the five recent crystallized CXCR4 structures along with a CXCR4 rhodopsin-based homology model. All CXCR4 Protein Data Bank (PDB) structures are co-crystallized with a small organic antagonist except structure 3OE0, which is co-crystallized with a cyclic peptide analog. Evaluation of the CXCR4 models was done by retrospective docking-based VS using a test set of 248 known CXCR4 inhibitors from 4 different chemotype families and 4696 different presumed inactives. The performance of the docking protocol using the five different protein structures was assessed in terms of pose prediction and hits detection using 12 different docking scoring functions and a scoring function with rescoring. Results show that 3OE6 structure achieves the highest docking-based performance with an average area under the curve (aAUC) of 0.84 and an average enrichment factor (aEF) of 11.7 at 1% of decoys screened. CXCR4 rhodopsin-like homology model performs comparable to the crystallized structures in the 1% of database screened. Moreover, a detailed analysis of the retrospective docking results using the CXCR4 homology model in Discovery Studio allows us to hypothesize the existence of multiple binding sub-sites in CXCR4 binding pocket.

FRED and HYBRID docking performance on standardized datasets

The docking performance of the FRED and HYBRID programs are evaluated on two standardized datasets from the Docking and Scoring Symposium of the ACS Spring 2011 national meeting. The evaluation includes cognate docking and virtual screening performance. FRED docks 70% of the structures to within 2Å in the cognate docking test. In the virtual screening test, FRED is found to have a mean AUC of 0.75. The HYBRID program uses a modified version of FRED's algorithm that uses both ligand- and structure-based information to dock molecules, which increases its mean AUC to 0.78. HYBRID can also implicitly account for protein flexibility by making use of multiple crystal structures. Using multiple crystal structures improves HYBRID's performance (mean AUC 0.80) with a negligible increase in docking time (~15%).

Surflex-Dock: Docking benchmarks and real-world application

Benchmarks for molecular docking have historically focused on re-docking the cognate ligand of a well-determined protein-ligand complex to measure geometric pose prediction accuracy, and measurement of virtual screening performance has been focused on increasingly large and diverse sets of target protein structures, cognate ligands, and various types of decoy sets. Here, pose prediction is reported on the Astex Diverse set of 85 protein ligand complexes, and virtual screening performance is reported on the DUD set of 40 protein targets. In both cases, prepared structures of targets and ligands were provided by symposium organizers. The re-prepared data sets yielded results not significantly different than previous reports of Surflex-Dock on the two benchmarks. Minor changes to protein coordinates resulting from complex pre-optimization had large effects on observed performance, highlighting the limitations of cognate ligand re-docking for pose prediction assessment. Docking protocols developed for cross-docking, which address protein flexibility and produce discrete families of predicted poses, produced substantially better performance for pose prediction. Performance on virtual screening performance was shown to benefit by employing and combining multiple screening methods: docking, 2D molecular similarity, and 3D molecular similarity. In addition, use of multiple protein conformations significantly improved screening enrichment.

The assessment of computationally derived protein ensembles in protein-ligand docking

The inclusion of receptor flexibility in protein-ligand docking experiments has become a major research interest in drug discovery [1,2]. One of the possible methods applied is the use of multiple discrete protein conformations, so called ensemble docking [3,4]. With computational techniques like Molecular Dynamics (MD) a large number of different conformations can be generated, not all of which can or should be included in the docking or virtual screening process [5]. The question arises if and how suitable protein conformations can be selected systematically a priori based on quantifiable conformational features. For neuraminidase and cyclin-dependent kinase II (CDK2), snapshots of MD simulation trajectories have been clustered based on different structural properties using a variety of clustering methods. To establish a possible correlation between docking performance and target conformational attributes the clustered snapshots have been subjected to extensive self- and cross-docking experiments as well as virtual screening using the GOLD docking programme. It is shown that conformationally similar snapshots do not necessarily result in a similar docking or virtual screening performance. The selection of the particular structural property on which to base the clustering appears to be the essential problem.

Ligand-Based Virtual Screening Approach Using a New Scoring Function

In this study, we aimed to develop a new ligand-based virtual screening approach using an effective shape-overlapping procedure and a more robust scoring function (denoted by the HWZ score for convenience). The HWZ score-based virtual screening approach was tested against the compounds for 40 protein targets available in the Database of Useful Decoys (DUD; dud.docking.org/jahn/ ), and the virtual screening performance was evaluated in terms of the area under the receiver operator characteristic (ROC) curve (AUC), enrichment factor (EF), and hit rate (HR), demonstrating an improved overall performance compared to other popularly used approaches examined. In particular, the HWZ score-based virtual screening led to an average AUC value of 0.84 ± 0.02 (95% confidence interval) for the 40 targets. The average HR values at the top 1% and 10% of the active compounds for the 40 targets were 46.3% ± 6.7% and 59.2% ± 4.7%, respectively. In addition, the performance of the HWZ score-based virtual screening approach is less sensitive to the choice of the target.

Pose prediction and virtual screening performance of GOLD scoring functions in a standardized test

The performance of all four GOLD scoring functions has been evaluated for pose prediction and virtual screening under the standardized conditions of the comparative docking and scoring experiment reported in this Edition. Excellent pose prediction and good virtual screening performance was demonstrated using unmodified protein models and default parameter settings. The best performing scoring function for both pose prediction and virtual screening was demonstrated to be the recently introduced scoring function ChemPLP. We conclude that existing docking programs already perform close to optimally in the cognate pose prediction experiments currently carried out and that more stringent pose prediction tests should be used in the future. These should employ cross-docking sets. Evaluation of virtual screening performance remains problematic and much remains to be done to improve the usefulness of publically available active and decoy sets for virtual screening. Finally we suggest that, for certain target/scoring function combinations, good enrichment may sometimes be a consequence of 2D property recognition rather than a modelling of the correct 3D interactions.

Computational fragment-based screening using RosettaLigand: the SAMPL3 challenge

SAMPL3 fragment based virtual screening challenge provides a valuable opportunity for researchers to test their programs, methods and screening protocols in a blind testing environment. We participated in SAMPL3 challenge and evaluated our virtual fragment screening protocol, which involves RosettaLigand as the core component by screening a 500 fragments Maybridge library against bovine pancreatic trypsin. Our study reaffirmed that the real test for any virtual screening approach would be in a blind testing environment. The analyses presented in this paper also showed that virtual screening performance can be improved, if a set of known active compounds is available and parameters and methods that yield better enrichment are selected. Our study also highlighted that to achieve accurate orientation and conformation of ligands within a binding site, selecting an appropriate method to calculate partial charges is important. Another finding is that using multiple receptor ensembles in docking does not always yield better enrichment than individual receptors. On the basis of our results and retrospective analyses from SAMPL3 fragment screening challenge we anticipate that chances of success in a fragment screening process could be increased significantly with careful selection of receptor structures, protein flexibility, sufficient conformational sampling within binding pocket and accurate assignment of ligand and protein partial charges.

Virtual Screening Data Fusion Using Both Structure- and Ligand-Based Methods

Virtual screening is widely applied in drug discovery, and significant effort has been put into improving current methods. In this study, we have evaluated the performance of compound ranking in virtual screening using five different data fusion algorithms on a total of 16 data sets. The data were generated by docking, pharmacophore search, shape similarity, and electrostatic similarity, spanning both structure- and ligand-based methods. The algorithms used for data fusion were sum rank, rank vote, sum score, Pareto ranking, and parallel selection. None of the fusion methods require any prior knowledge or input other than the results from the single methods and, thus, are readily applicable. The results show that compound ranking using data fusion improves the performance and consistency of virtual screening compared to the single methods alone. The best performing data fusion algorithm was parallel selection, but both rank voting and Pareto ranking also have good performance.

Evaluation of docking performance in a blinded virtual screening of fragment-like trypsin inhibitors

In this study, we have "blindly" assessed the ability of several combinations of docking software and scoring functions to predict the binding of a fragment-like library of bovine trypsine inhibitors. The most suitable protocols (involving Gold software and GoldScore scoring function, with or without rescoring) were selected for this purpose using a training set of compounds with known biological activities. The selected virtual screening protocols provided good results with the SAMPL3-VS dataset, showing enrichment factors of about 10 for Top 20 compounds. This methodology should be useful in difficult cases of docking, with a special emphasis on the fragment-based virtual screening campaigns.

Cheminformatics Meets Molecular Mechanics: A Combined Application of Knowledge-Based Pose Scoring and Physical Force Field-Based Hit Scoring Functions Improves the Accuracy of Structure-Based Virtual Screening

Poor performance of scoring functions is a well-known bottleneck in structure-based virtual screening (VS), which is most frequently manifested in the scoring functions' inability to discriminate between true ligands vs known nonbinders (therefore designated as binding decoys). This deficiency leads to a large number of false positive hits resulting from VS. We have hypothesized that filtering out or penalizing docking poses recognized as non-native (i.e., pose decoys) should improve the performance of VS in terms of improved identification of true binders. Using several concepts from the field of cheminformatics, we have developed a novel approach to identifying pose decoys from an ensemble of poses generated by computational docking procedures. We demonstrate that the use of target-specific pose (scoring) filter in combination with a physical force field-based scoring function (MedusaScore) leads to significant improvement of hit rates in VS studies for 12 of the 13 benchmark sets from the clustered version of the Database of Useful Decoys (DUD). This new hybrid scoring function outperforms several conventional structure-based scoring functions, including XSCORE::HMSCORE, ChemScore, PLP, and Chemgauss3, in 6 out of 13 data sets at early stage of VS (up 1% decoys of the screening database). We compare our hybrid method with several novel VS methods that were recently reported to have good performances on the same DUD data sets. We find that the retrieved ligands using our method are chemically more diverse in comparison with two ligand-based methods (FieldScreen and FLAP::LBX). We also compare our method with FLAP::RBLB, a high-performance VS method that also utilizes both the receptor and the cognate ligand structures. Interestingly, we find that the top ligands retrieved using our method are highly complementary to those retrieved using FLAP::RBLB, hinting effective directions for best VS applications. We suggest that this integrative VS approach combining cheminformatics and molecular mechanics methodologies may be applied to a broad variety of protein targets to improve the outcome of structure-based drug discovery studies.

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.

SHAFTS: A Hybrid Approach for 3D Molecular Similarity Calculation. 1. Method and Assessment of Virtual Screening

We developed a novel approach called SHAFTS (SHApe-FeaTure Similarity) for 3D molecular similarity calculation and ligand-based virtual screening. SHAFTS adopts a hybrid similarity metric combined with molecular shape and colored (labeled) chemistry groups annotated by pharmacophore features for 3D similarity calculation and ranking, which is designed to integrate the strength of pharmacophore matching and volumetric overlay approaches. A feature triplet hashing method is used for fast molecular alignment poses enumeration, and the optimal superposition between the target and the query molecules can be prioritized by calculating corresponding "hybrid similarities". SHAFTS is suitable for large-scale virtual screening with single or multiple bioactive compounds as the query "templates" regardless of whether corresponding experimentally determined conformations are available. Two public test sets (DUD and Jain's sets) including active and decoy molecules from a panel of useful drug targets were adopted to evaluate the virtual screening performance. SHAFTS outperformed several other widely used virtual screening methods in terms of enrichment of known active compounds as well as novel chemotypes, thereby indicating its robustness in hit compounds identification and potential of scaffold hopping in virtual screening.

VSDK: Virtual screening of small molecules using AutoDock Vina on Windows platform

Screening of ligand molecules to target proteins using computer-aided docking is a critical step in rational drug discovery. Based on this circumstance, we attempted to develop a virtual screening application system, named VSDK Virtual Screening by Docking, which can function under the Windows platform. This is a user-friendly, flexible, and versatile tool which can be used by users who are familiar with Windows OS. The virtual screening performance was tested for an arbitrarily-selected receptor, FGFR tyrosine kinase (pdb code: 1agw), by using ligands downloaded from ZINC database with its grid size of x,y,z = 30,30,30 and run number of 10. It took 90 minutes for 100 molecules for this virtual screening. VSDK is freely available at the designated URL, and a simplified manual can be downloaded from VSDK home page. This tool will have a more challenging scope and achievement as the computer speed and accuracy are increased and secured in the future. The database is available for free at http://www.pharm.kobegakuin.ac.jp/˜akaho/english_top.html.

Systematic Exploitation of Multiple Receptor Conformations for Virtual Ligand Screening

The role of virtual ligand screening in modern drug discovery is to mine large chemical collections and to prioritize for experimental testing a comparatively small and diverse set of compounds with expected activity against a target. Several studies have pointed out that the performance of virtual ligand screening can be improved by taking into account receptor flexibility. Here, we systematically assess how multiple crystallographic receptor conformations, a powerful way of discretely representing protein plasticity, can be exploited in screening protocols to separate binders from non-binders. Our analyses encompass 36 targets of pharmaceutical relevance and are based on actual molecules with reported activity against those targets. The results suggest that an ensemble receptor-based protocol displays a stronger discriminating power between active and inactive molecules as compared to its standard single rigid receptor counterpart. Moreover, such a protocol can be engineered not only to enrich a higher number of active compounds, but also to enhance their chemical diversity. Finally, some clear indications can be gathered on how to select a subset of receptor conformations that is most likely to provide the best performance in a real life scenario.

Evaluation of different machine learning methods for ligand-based virtual screening

In silico High Throughput Screening of large compound databases has become increasingly popular technology of finding valuable drug candidates, by applying a wide range of computational methods, such as machine learning [1]. In recent years, many comparative studies of different machine learning methods performance in ligand-based virtual screening have been reported [2,3]. In order to extend these studies, we have evaluated over 60 different machine learning methods, such as: support vector machines (with and without parameter optimization), naive Bayesian, decision trees, random forest, meta-classifiers (boosting, bagging, grading) and many others. All calculations were performed using a collection of machine learning algorithms for data mining implemented in WEKA package [4]. Additionally, for each of the method, we have examined the influence of different type of fingerprints, the size of training sets and attribute selection methods on the rate of active recall and precision of selection. Our internal database of known 5-HT7 antagonists has been used to build training and testing sets. It was found that there is no machine learning approach that consistently provides the best results but some of them are very stable and can be applied universally.

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.

Pharmacophore alignment search tool: Influence of the third dimension on text‐based similarity searching

Previously (Hähnke et al., J Comput Chem 2010, 31, 2810) we introduced the concept of nonlinear dimensionality reduction for canonization of two-dimensional layouts of molecular graphs as foundation for text-based similarity searching using our Pharmacophore Alignment Search Tool (PhAST), a ligand-based virtual screening method. Here we apply these methods to three-dimensional molecular conformations and investigate the impact of these additional degrees of freedom on virtual screening performance and assess differences in ranking behavior. Best-performing variants of PhAST are compared with 16 state-of-the-art screening methods with respect to significance estimates for differences in screening performance. We show that PhAST sorts new chemotypes on early ranks without sacrificing overall screening performance. We succeeded in combining PhAST with other virtual screening techniques by rank-based data fusion, significantly improving screening capabilities. We also present a parameterization of double dynamic programming for the problem of small molecule comparison, which allows for the calculation of structural similarity between compounds based on one-dimensional representations, opening the door to a holistic approach to molecule comparison based on textual representations.