Docking pose selection by interaction pattern graph similarity: application to the D3R grand challenge 2015

One class classification for the detection of β2 adrenergic receptor agonists using single-ligand dynamic interaction data

G protein-coupled receptors are involved in many biological processes, relaying the extracellular signal inside the cell. Signaling is regulated by the interactions between receptors and their ligands, it can be stimulated by agonists, or inhibited by antagonists or inverse agonists. The development of a new drug targeting a member of this family requires to take into account the pharmacological profile of the designed ligands in order to elicit the desired response. The structure-based virtual screening of chemical libraries may prioritize a specific class of ligands by combining docking results and ligand binding information provided by crystallographic structures. The performance of the method depends on the relevance of the structural data, in particular the conformation of the targeted site, the binding mode of the reference ligand, and the approach used to compare the interactions formed by the docked ligand with those formed by the reference ligand in the crystallographic structure. Here, we propose a new method based on the conformational dynamics of a single protein–ligand reference complex to improve the biased selection of ligands with specific pharmacological properties in a structure-based virtual screening exercise. Interactions patterns between a reference agonist and the receptor, here exemplified on the β2 adrenergic receptor, were extracted from molecular dynamics simulations of the agonist/receptor complex and encoded in graphs used to train a one-class machine learning classifier. Different conditions were tested: low to high affinity agonists, varying simulation duration, considering or ignoring hydrophobic contacts, and tuning of the classifier parametrization. The best models applied to post-process raw data from retrospective virtual screening obtained by docking of test libraries effectively filtered out irrelevant poses, discarding inactive and non-agonist ligands while identifying agonists. Taken together, our results suggest that consistency of the binding mode during the simulation is a key to the success of the method.

Mebendazole's Conformational Space and Its Predicted Binding to Human Heat-Shock Protein 90

Recent experimental evidence suggests that mebendazole, a popular antiparasitic drug, binds to heat shock protein 90 (Hsp90) and inhibits acute myeloid leukemia cell growth. In this study we use quantum mechanics (QM), molecular similarity, and molecular dynamics (MD) calculations to predict possible binding poses of mebendazole to the adenosine triphosphate (ATP) binding site of Hsp90. Extensive conformational searches and minimization of the five mebendazole tautomers using the MP2/aug-cc-pVTZ theory level resulted in 152 minima. Mebendazole-Hsp90 complex models were subsequently created using the QM optimized conformations and protein coordinates obtained from experimental crystal structures that were chosen through similarity calculations. Nine different poses were identified from a total of 600 ns of explicit solvent, all-atom MD simulations using two different force fields. All simulations support the hypothesis that mebendazole is able to bind to the ATP binding site of Hsp90.

True Accuracy of Fast Scoring Functions to Predict High-Throughput Screening Data from Docking Poses: The Simpler the Better

Hundreds of fast scoring functions have been developed over the last 20 years to predict binding free energies from three-dimensional structures of protein-ligand complexes. Despite numerous statistical promises, we believe that none of them has been properly validated for daily prospective high-throughput virtual screening studies, mostly because in silico screening challenges usually employ artificially built and biased datasets. We here carry out a fully unbiased evaluation of four scoring functions (Pafnucy, Δ_vinaRF₂₀, IFP, and GRIM) on an in-house developed data collection of experimental high-confidence screening data (LIT-PCBA) covering about 3 million data points on 15 diverse pharmaceutical targets. All four scoring functions were applied to rescore the docking poses of LIT-PCBA compounds in conditions mimicking exactly standard drug discovery scenarios and were compared in terms of propensity to enrich true binders in the top 1%-ranked hit lists. Interestingly, rescoring based on simple interaction fingerprints or interaction graphs outperforms state-of-the-art machine learning and deep learning scoring functions in most of the cases. The current study notably highlights the strong tendency of deep learning methods to predict affinity values within a very narrow range centered on the mean value of samples used for training. Moreover, it suggests that knowledge of pre-existing binding modes is the key to detecting the most potent binders.

Potential repurposing of four FDA approved compounds with antiplasmodial activity identified through proteome scale computational drug discovery and in vitro assay

Malaria elimination can benefit from time and cost-efficient approaches for antimalarials such as drug repurposing. In this work, 796 DrugBank compounds were screened against 36 Plasmodium falciparum targets using QuickVina-W. Hits were selected after rescoring using GRaph Interaction Matching (GRIM) and ligand efficiency metrics: surface efficiency index (SEI), binding efficiency index (BEI) and lipophilic efficiency (LipE). They were further evaluated in Molecular dynamics (MD). Twenty-five protein-ligand complexes were finally retained from the 28,656 (36 × 796) dockings. Hit GRIM scores (0.58 to 0.78) showed their molecular interaction similarity to co-crystallized ligands. Minimum LipE (3), SEI (23) and BEI (7) were in at least acceptable thresholds for hits. Binding energies ranged from -6 to -11 kcal/mol. Ligands showed stability in MD simulation with good hydrogen bonding and favorable protein-ligand interactions energy (the poorest being -140.12 kcal/mol). In vitro testing showed 4 active compounds with two having IC50 values in the single-digit μM range.

Using the Semiempirical Quantum Mechanics in Improving the Molecular Docking: A Case Study with CDK2

In this study, we use some modified semiempirical quantum mechanics (SQM) methods for improving the molecular docking process. To this end, the three popular SQM Hamiltonians, PM6, PM6-D3H4X, and PM7 are employed for geometry optimization of some binding modes of ligands docked into the human cyclin-dependent kinase 2 (CDK2) by two widely used docking tools, AutoDock and AutoDock Vina. The results were analyzed with two different evaluation metrics: the symmetry-corrected heavy-atom RMSD and the fraction of recovered ligand-protein contacts. It is shown that the evaluation of the fraction of recovered contacts is more useful to measure the similarity between two structures when interacting with a protein. It was also found that AutoDock is more successful than AutoDock Vina in producing the correct ligand poses (RMSD≤2.0 Å) and ranking of the poses. It is also demonstrated that the ligand optimization at the SQM level improves the docking results and the SQM structures have a significantly better fit to the observed crystal structures. Finally, the SQM optimizations reduce the number of close contacts in the docking poses and successfully remove most of the clash or bad contacts between ligand and protein.

Combining Docking Pose Rank and Structure with Deep Learning Improves Protein-Ligand Binding Mode Prediction over a Baseline Docking Approach

We present a simple, modular graph-based convolutional neural network that takes structural information from protein-ligand complexes as input to generate models for activity and binding mode prediction. Complex structures are generated by a standard docking procedure and fed into a dual-graph architecture that includes separate subnetworks for the ligand bonded topology and the ligand-protein contact map. Recent work has indicated that data set bias drives many past promising results derived from combining deep learning and docking. Our dual-graph network allows contributions from ligand identity that give rise to such biases to be distinguished from effects of protein-ligand interactions on classification. We show that our neural network is capable of learning from protein structural information when, as in the case of binding mode prediction, an unbiased data set is constructed. We next develop a deep learning model for binding mode prediction that uses docking ranking as input in combination with docking structures. This strategy mirrors past consensus models and outperforms a baseline docking program (AutoDock Vina) in a variety of tests, including on cross-docking data sets that mimic real-world docking use cases. Furthermore, the magnitudes of network predictions serve as reliable measures of model confidence.

Getting Docking into Shape Using Negative Image-Based Rescoring

The failure of default scoring functions to ensure virtual screening enrichment is a persistent problem for the molecular docking algorithms used in structure-based drug discovery. To remedy this problem, elaborate rescoring and postprocessing schemes have been developed with a varying degree of success, specificity, and cost. The negative image-based rescoring (R-NiB) has been shown to improve the flexible docking performance markedly with a variety of drug targets. The yield improvement is achieved by comparing the alternative docking poses against the negative image of the target protein's ligand-binding cavity. In other words, the shape and electrostatics of the binding pocket is directly used in the similarity comparison to rank the explicit docking poses. Here, the PANTHER/ShaEP-based R-NiB methodology is tested with six popular docking softwares, including GLIDE, PLANTS, GOLD, DOCK, AUTODOCK, and AUTODOCK VINA, using five validated benchmark sets. Overall, the results indicate that R-NiB outperforms the default docking scoring consistently and inexpensively, demonstrating that the methodology is ready for wide-scale virtual screening usage.

Local Interaction Density (LID), a Fast and Efficient Tool to Prioritize Docking Poses

Ligand docking at a protein site can be improved by prioritizing poses by similarity to validated binding modes found in the crystal structures of ligand/protein complexes. The interactions formed in the predicted model are searched in each of the reference 3D structures, taken individually. We propose to merge the information provided by all references, creating a single representation of all known binding modes. The method is called LID, an acronym for Local Interaction Density. LID was benchmarked in a pose prediction exercise on 19 proteins and 1382 ligands using PLANTS as docking software. It was also tested in a virtual screening challenge on eight proteins, with a dataset of 140,000 compounds from DUD-E and PubChem. LID significantly improved the performance of the docking program in both pose prediction and virtual screening. The gain is comparable to that obtained with a rescoring approach based on the individual comparison of reference binding modes (the GRIM method). Importantly, LID is effective with a small number of references. LID calculation time is negligible compared to the docking time.

A benchmark driven guide to binding site comparison: An exhaustive evaluation using tailor-made data sets (ProSPECCTs)

The automated comparison of protein-ligand binding sites provides useful insights into yet unexplored site similarities. Various stages of computational and chemical biology research can benefit from this knowledge. The search for putative off-targets and the establishment of polypharmacological effects by comparing binding sites led to promising results for numerous projects. Although many cavity comparison methods are available, a comprehensive analysis to guide the choice of a tool for a specific application is wanting. Moreover, the broad variety of binding site modeling approaches, comparison algorithms, and scoring metrics impedes this choice. Herein, we aim to elucidate strengths and weaknesses of binding site comparison methodologies. A detailed benchmark study is the only possibility to rationalize the selection of appropriate tools for different scenarios. Specific evaluation data sets were developed to shed light on multiple aspects of binding site comparison. An assembly of all applied benchmark sets (ProSPECCTs–Protein Site Pairs for the Evaluation of Cavity Comparison Tools) is made available for the evaluation and optimization of further and still emerging methods. The results indicate the importance of such analyses to facilitate the choice of a methodology that complies with the requirements of a specific scientific challenge.

Binding site similarities are useful in the context of promiscuity prediction, drug repurposing, the analysis of protein-ligand and protein-protein complexes, function prediction, and further fields of general interest in chemical biology and biochemistry. Many years of research have led to the development of a multitude of methods for binding site analysis and comparison. On the one hand, their availability supports research. On the other hand, the huge number of methods hampers the efficient selection of a specific tool. Our research is dedicated to the analysis of different cavity comparison tools. We use several binding site data sets to establish guidelines which can be applied to ensure a successful application of comparison methods by circumventing potential pitfalls.

PLHINT: A knowledge-driven computational approach based on the intermolecular H bond interactions at the protein-ligand interface from docking solutions

The tendency of docking scoring functions to generate crystal close conformations of ligands bound to protein structures face limitations in not reproducing the exact crystal intermolecular contacts in dock poses. Intermolecular H bond contacts enumerated at the protein-docked ligand interface can be used to train scoring models and improve virtual screening performance. There is a need to incorporate additional knowledge of protein-ligand H bond contacts in extension to crystal contacts from docking solutions within the reproducibility efficiency of the docking program. A computational approach PLHINT (Protein-ligand H bond interaction pattern) is presented here which extracts intermolecular H bond interactions from native-like docked ligand poses, transform into the scoring scheme and apply over the virtual screening results of database molecules. The basic premise of the PLHINT approach is to score the most observed H bond patterns with the high score to achieve high recovery rates. Tested on ten diverse DUD-E benchmark datasets, the approach has demonstrated better overall performance and ligand enrichment competency over virtual screening results generated by three genetic algorithm-based docking programs viz. AutoDock Vina, FlexAID and PLANTS. Furthermore, the approach has successfully recovered the poor and random virtual screening results with better enrichments.

IChem: A Versatile Toolkit for Detecting, Comparing, and Predicting Protein-Ligand Interactions

Structure-based ligand design requires an exact description of the topology of molecular entities under scrutiny. IChem is a software package that reflects the many contributions of our research group in this area over the last decade. It facilitates and automates many tasks (e.g., ligand/cofactor atom typing, identification of key water molecules) usually left to the modeler's choice. It therefore permits the detection of molecular interactions between two molecules in a very precise and flexible manner. Moreover, IChem enables the conversion of intricate three-dimensional (3D) molecular objects into simple representations (fingerprints, graphs) that facilitate knowledge acquisition at very high throughput. The toolkit is an ideal companion for setting up and performing many structure-based design computations.

Ranking docking poses by graph matching of protein-ligand interactions: lessons learned from the D3R Grand Challenge 2

A novel docking challenge has been set by the Drug Design Data Resource (D3R) in order to predict the pose and affinity ranking of a set of Farnesoid X receptor (FXR) agonists, prior to the public release of their bound X-ray structures and potencies. In a first phase, 36 agonists were docked to 26 Protein Data Bank (PDB) structures of the FXR receptor, and next rescored using the in-house developed GRIM method. GRIM aligns protein-ligand interaction patterns of docked poses to those of available PDB templates for the target protein, and rescore poses by a graph matching method. In agreement with results obtained during the previous 2015 docking challenge, we clearly show that GRIM rescoring improves the overall quality of top-ranked poses by prioritizing interaction patterns already visited in the PDB. Importantly, this challenge enables us to refine the applicability domain of the method by better defining the conditions of its success. We notably show that rescoring apolar ligands in hydrophobic pockets leads to frequent GRIM failures. In the second phase, 102 FXR agonists were ranked by decreasing affinity according to the Gibbs free energy of the corresponding GRIM-selected poses, computed by the HYDE scoring function. Interestingly, this fast and simple rescoring scheme provided the third most accurate ranking method among 57 contributions. Although the obtained ranking is still unsuitable for hit to lead optimization, the GRIM-HYDE scoring scheme is accurate and fast enough to post-process virtual screening data.