Scoring functions for protein-ligand interactions: a critical perspective

Discovery of a Hidden Trypanosoma cruzi Spermidine Synthase Binding Site and Inhibitors through In Silico, In Vitro, and X-ray Crystallography

In drug discovery research, the selection of promising binding sites and understanding the binding mode of compounds are crucial fundamental studies. The current understanding of the proteins-ligand binding model extends beyond the simple lock and key model to include the induced-fit model, which alters the conformation to match the shape of the ligand, and the pre-existing equilibrium model, selectively binding structures with high binding affinity from a diverse ensemble of proteins. Although methods for detecting target protein binding sites and virtual screening techniques using docking simulation are well-established, with numerous studies reported, they only consider a very limited number of structures in the diverse ensemble of proteins, as these methods are applied to a single structure. Molecular dynamics (MD) simulation is a method for predicting protein dynamics and can detect potential ensembles of protein binding sites and hidden sites unobservable in a single-point structure. In this study, to demonstrate the utility of virtual screening with protein dynamics, MD simulations were performed on Trypanosoma cruzi spermidine synthase to obtain an ensemble of dominant binding sites with a high probability of existence. The structure of the binding site obtained through MD simulation revealed pockets in addition to the active site that was present in the initial structure. Using the obtained binding site structures, virtual screening of 4.8 million compounds by docking simulation, in vitro assays, and X-ray analysis was conducted, successfully identifying two hit compounds.

Leveraging scaffold information to predict protein-ligand binding affinity with an empirical graph neural network

Protein-ligand binding affinity prediction is an important task in structural bioinformatics for drug discovery and design. Although various scoring functions (SFs) have been proposed, it remains challenging to accurately evaluate the binding affinity of a protein-ligand complex with the known bound structure because of the potential preference of scoring system. In recent years, deep learning (DL) techniques have been applied to SFs without sophisticated feature engineering. Nevertheless, existing methods cannot model the differential contribution of atoms in various regions of proteins, and the relationship between atom properties and intermolecular distance is also not fully explored. We propose a novel empirical graph neural network for accurate protein-ligand binding affinity prediction (EGNA). Graphs of protein, ligand and their interactions are constructed based on different regions of each bound complex. Proteins and ligands are effectively represented by graph convolutional layers, enabling the EGNA to capture interaction patterns precisely by simulating empirical SFs. The contributions of different factors on binding affinity can thus be transparently investigated. EGNA is compared with the state-of-the-art machine learning-based SFs on two widely used benchmark data sets. The results demonstrate the superiority of EGNA and its good generalization capability.

AlteQ: a new complementarity principle-centered method for the evaluation of docking poses

Molecular docking is the most popular and widely used method for identifying novel molecules against a target of interest. However, docking procedures and their validation are still under intense development. In the present investigation, we evaluate a quantum free-orbital AlteQ method for evaluating docking complexes generated by taking EGFR complexes as an example. The AlteQ method calculates the electron density using Slater's type atomic contributions in the interspace between the receptor and the ligand. Since the interactions are determined by the overlap of electron clouds, they follow the complementarity principle, and an equation can be obtained that describes these interactions. The AlteQ method evaluates the quality of the interaction between the receptor and the ligand, how complementary the interactions are, and due to this, it is used to reject less realistic structures obtained by docking methods. Here, three different equations were used to determine the quality of the interactions in experimental complexes and docked complexes obtained using AutoDock Vina and AutoDock 4.2.6.Communicated by Ramaswamy H. Sarma.

The Mechanisms Underlying the Pharmacological Effects of GuiPi Decoction on Major Depressive Disorder based on Network Pharmacology and Molecular Docking

BACKGROUND AND AIM

Major Depressive Disorder (MDD) is a common affective disorder. GuiPi decoction (GPD) is used to treat depression in China, Japan, and Korea. However, its effective ingredients and antidepressant mechanisms remain unclear. We attempted to reveal the potential mechanisms of GPD in the treatment of MDD by network pharmacology and molecular docking. In addition, we conducted an enzymatic activity assay to validate the results of molecular docking.

METHODS

GPD-related compounds and targets, and MDD-related targets were retrieved from databases and literature. The herb-compound-target network was constructed by Cytoscape. The protein- protein interaction network was built using the STRING database to find key targets of GPD on MDD. Enrichment analysis of shared targets was analyzed by MetaCore database to obtain the potential pathway and biological process of GPD on MDD. The main active compounds treating MDD were screened by molecular docking. The PDE4s inhibitors were screened and verified by an enzyme activity assay.

RESULTS

GPD contained 1222 ingredients and 190 potential targets for anti-MDD. Possible biological processes regulated by GPD were neurophysiological processes, blood vessel morphogenesis, Camp Responsive Element Modulator (CREM) pathway, and Androgen Receptor (AR) signaling crosstalk in MDD. Potential pathways in MDD associated with GPD include neurotransmission, cell differentiation, androgen signaling, and estrogen signaling. Fumarine, m-cresol, quercetin, betasitosterol, fumarine, taraxasterol, and lupeol in GPD may be the targets of SLC6A4, monoamine oxidase A (MAOA), DRD2, OPRM1, HTR3A, Albumin (ALB), and NTRK1, respectively. The IC50 values of trifolin targeting Phosphodiesterase (PDE) 4A and girinimbine targeting PDE4B1 were 73.79 μM and 31.86 μM, respectively. The IC50 values of girinimbine and benzo[a]carbazole on PDE4B2 were 51.62 μM and 94.61 μM, respectively.

CONCLUSION

Different compounds in GPD may target the same protein, and the same component in GPD can target multiple targets. These results suggest that the effects of GPD on MDD are holistic and systematic, unlike the pattern of one drug-one target.

On drug discovery against infectious diseases and academic medicinal chemistry contributions

This perspective is an attempt to document the problems that medicinal chemists are facing in drug discovery. It is also trying to identify relevant/possible, research areas in which academics can have an impact and should thus be the subject of grant calls. Accordingly, it describes how hit discovery happens, how compounds to be screened are selected from available chemicals and the possible reasons for the recurrent paucity of useful/exploitable results reported. This is followed by the successful hit to lead stories leading to recent and original antibacterials which are, or about to be, used in human medicine. Then, illustrated considerations and suggestions are made on the possible inputs of academic medicinal chemists. This starts with the observation that discovering a “good” hit in the course of a screening campaign still rely on a lot of luck – which is within the reach of academics –, that the hit to lead process requires a lot of chemistry and that if public–private partnerships can be important throughout these stages, they are absolute requirements for clinical trials. Concerning suggestions to improve the current hit success rate, one academic input in organic chemistry would be to identify new and pertinent chemical space, design synthetic accesses to reach these and prepare the corresponding chemical libraries. Concerning hit to lead programs on a given target, if no new hits are available, previously reported leads along with new structural data can be pertinent starting points to design, prepare and assay original analogues. In conclusion, this text is an actual plea illustrating that, in many countries, academic research in medicinal chemistry should be more funded, especially in the therapeutic area neglected by the industry. At the least, such funds would provide the intensive to secure series of hopefully relevant chemical entities which appears to often lack when considering the results of academic as well as industrial screening campaigns.

3pHLA-score improves structure-based peptide-HLA binding affinity prediction

Binding of peptides to Human Leukocyte Antigen (HLA) receptors is a prerequisite for triggering immune response. Estimating peptide-HLA (pHLA) binding is crucial for peptide vaccine target identification and epitope discovery pipelines. Computational methods for binding affinity prediction can accelerate these pipelines. Currently, most of those computational methods rely exclusively on sequence-based data, which leads to inherent limitations. Recent studies have shown that structure-based data can address some of these limitations. In this work we propose a novel machine learning (ML) structure-based protocol to predict binding affinity of peptides to HLA receptors. For that, we engineer the input features for ML models by decoupling energy contributions at different residue positions in peptides, which leads to our novel per-peptide-position protocol. Using Rosetta's ref2015 scoring function as a baseline we use this protocol to develop 3pHLA-score. Our per-peptide-position protocol outperforms the standard training protocol and leads to an increase from 0.82 to 0.99 of the area under the precision-recall curve. 3pHLA-score outperforms widely used scoring functions (AutoDock4, Vina, Dope, Vinardo, FoldX, GradDock) in a structural virtual screening task. Overall, this work brings structure-based methods one step closer to epitope discovery pipelines and could help advance the development of cancer and viral vaccines.

Assessment of CYP2C9 Structural Models for Site of Metabolism Prediction

Structure-based prediction of a compound's potential sites of metabolism (SOMs) mediated by cytochromes P450 (CYPs) is highly advantageous in the early stage of drug discovery. However, the accuracy of the SOMs prediction can be influenced by several factors. CYP2C9 is one of the major drug-metabolizing enzymes in humans and is responsible for the metabolism of ∼13 % of clinically used drugs. In this study, we systematically evaluated the effects of protein crystal structure models, scoring functions, heme forms, conserved active-site water molecules, and protein flexibility on SOMs prediction of CYP2C9 substrates. Our results demonstrated that, on average, ChemScore and GlideScore outperformed four other scoring functions: Vina, GoldScore, ChemPLP, and ASP. The performance of the crystal structure models with pentacoordinated heme was generally superior to that of the hexacoordinated iron-oxo heme (referred to as Compound I) models. Inclusion of the conserved active-site water molecule improved the prediction accuracy of GlideScore, but reduced the accuracy of ChemScore. In addition, the effect of the conserved water on SOMs prediction was found to be dependent on the receptor model and the substrate. We further found that one of snapshots from molecular dynamics simulations on the apo form can improve the prediction accuracy when compared to the crystal structural model.

Decision Making in Structure-Based Drug Discovery: Visual Inspection of Docking Results

Molecular docking is a computational method widely used in drug discovery. Due to the inherent inaccuracies of molecular docking, visual inspection of binding modes is a crucial routine in the decision making process of computational medicinal chemists. Despite its apparent importance for medicinal chemistry projects, guidelines for the visual docking pose assessment have been hardly discussed in the literature. Here, we review the medicinal chemistry literature with the aim of identifying consistent principles for visual inspection, highlighting cases of its successful application, and discussing its limitations. In this context, we conducted a survey reaching experts in both academia and the pharmaceutical industry, which also included a challenge to distinguish native from incorrect poses. We were able to collect 93 expert opinions that offer valuable insights into visually supported decision-making processes. This perspective shall motivate discussions among experienced computational medicinal chemists and guide young scientists new to the field to stratify their compounds.

A recursive molecular docking coupled with energy-based pose-rescoring and MD simulations to identify hsGC βH-NOX allosteric modulators for cardiovascular dysfunctions

Modulating the activity of human soluble guanylate cyclase (hsGC) through allosteric regulation of the βH-NOX domain has been considered as an immediate treatment for cardiovascular disorder (CVDs). Currently available βH-NOX domain-specific agonists including cinaciguat are unable to deal with the conundrum raised due to oxidative stress in the case of CVDs and their associated comorbidities. Therefore, the idea of investigating novel compounds for allosteric regulation of hsGC activation has been rekindled to circumvent CVDs. Current study aims to identify novel βH-NOX domain-specific compounds that can selectively turn on sGC functions by modulating the conformational dynamics of the target protein. Through a comprehensive computational drug-discovery approach, we first executed a target-based performance assessment of multiple docking (PLANTS, QVina, LeDock, Vinardo, Smina) scoring functions based on multiple performance metrices. QVina showed the highest capability of selecting true-positive ligands over false positives thus, used to screen 4.8 million ZINC15 compounds against βH-NOX domain. The docked ligands were further probed in terms of contact footprint and pose reassessment through clustering analysis and PLANTS docking, respectively. Subsequently, energy-based AMBER rescoring of top 100 low-energy complexes, per-residue energy decomposition analysis, and ADME-Tox analysis yielded the top three compounds i.e. ZINC000098973660, ZINC001354120371, and ZINC000096022607. The impact of three selected ligands on the internal structural dynamics of the βH-NOX domain was also investigated through molecular dynamics simulations. The study revealed potential electrostatic interactions for better conformational dialogue between βH-NOX domain and allosteric ligands that are critical for the activation of hsGC as compared to the reference compound.

Assessing multiple score functions in Rosetta for drug discovery

Rosetta is a computational software suite containing algorithms for a wide variety of macromolecular structure prediction and design tasks including small molecule protocols commonly used in drug discovery or enzyme design. Here, we benchmark RosettaLigand score functions and protocols in comparison to results of other software recently published in the Comparative Assessment of Score Functions (CASF-2016). The CASF-2016 benchmark covers a wide variety of tests including scoring and ranking multiple compounds against a target, ligand docking of a small molecule to a target, and virtual screening to extract binders from a compound library. Direct comparison to the score functions provided by CASF-2016 results shows that the original RosettaLigand score function ranks among the top software for scoring, ranking, docking and screening tests. Most notably, the RosettaLigand score function ranked 2/34 among other report score functions in CASF-2016. We additionally perform a ligand docking test with full sampling to mimic typical use cases. Despite improved performance of newer score functions in canonical protein structure prediction and design, we demonstrate here that more recent Rosetta score functions have reduced performance across all small molecule benchmarks. The tests described here have also been uploaded to the Rosetta scientific benchmarking server and will be run weekly to track performance as the code is continually being developed.

Identification of Novel Phyto-chemicals from Ocimum basilicum for the Treatment of Parkinson's Disease using In Silico Approach

BACKGROUND

Parkinson's disease is characterized by decreased level of dopaminergic neurotransmitters and this decrease is due to the degradation of dopamine by protein Monoamine Oxidase B (MAO-B). In order to treat Parkinson's disease, MAO-B should be inhibited.

OBJECTIVE

To find out the novel phytochemicals from plant Ocimum basilicum that can inhibit MAO-B by using the in silico methods.

METHODS

The data of chemical constituents from plant Ocimum basilicum was collected and inhibitory activity of these phytochemicals was then predicted by using the Structure-Based (SB) and Ligand-Based Virtual Screening (LBVS) methods. Molecular docking, one of the common Structure-Based Virtual Screening method, has been used during this search. Traditionally, molecular docking is used to predict the orientation and binding affinity of the ligand within the active site of the protein. Molegro Virtual Docker (MVD) software has been used for this purpose. On the other hand, Random Forest Model, one of the LBVS method, has also been used to predict the activity of these chemical constituents of Ocimum basilicum against the MAO-B.

RESULTS

During the docking studies, all the 108 compounds found in Ocimum basilicum were docked within the active site of MAO-B (PDB code: 4A79) out of which, 57 compounds successfully formed the hydrogen bond with tyr 435, a crucial amino acid for the biological activity of the enzyme. Rutin (-182.976 Kcal/mol), Luteolin (-163.171 Kcal/mol), Eriodictyol-7-O-glucoside (- 160.13 Kcal/mol), Rosmarinic acid (-133.484 Kcal/mol) and Isoquercitrin (-131.493 Kcal/mol) are among the top hits with the highest MolDock score along with hydrogen interaction with tyr 435. Using the RF model, ten compounds out of 108 chemical constituent of Ocimum basilicum were predicted to be active, Apigenin (1.0), Eriodictyol (1.0), Orientin (0.876), Kaempferol (0.8536), Luteolin (0.813953) and Rosmarinic-Acid (0.7738095) are predicted to be most active with the highest RF score.

CONCLUSION

The comparison of the two screening methods show that the ten compounds that were predicted to be active by the RF model, are also found in top hits of docking studies with the highest score. The top hits obtained during this study are predicted to be the inhibitor of MAO-B, thus, could be used further for the development of drugs for the treatment of Parkinson's disease (PD).

Improving the binding affinity estimations of protein-ligand complexes using machine-learning facilitated force field method

Scoring functions are routinely deployed in structure-based drug design to quantify the potential for protein-ligand (PL) complex formation. Here, we present a new scoring function Bappl+ that is designed to predict the binding affinities of non-metallo and metallo PL complexes. Bappl+ outperforms other state-of-the-art scoring functions, achieving a high Pearson correlation coefficient of up to ~ 0.76 with low standard deviations. The biggest contributors to the increased performance are the use of a machine-learning model and the enlarged training dataset. We have also evaluated the performance of Bappl+ on target-specific proteins, which highlighted the limitations of our function and provides a way for further improvements. We believe that Bappl+ methodology could prove valuable in ranking candidate molecules against a target metallo or non-metallo protein by reliably predicting their binding affinities, thus helping in the drug discovery process.

In silico repositioning of etidronate as a potential inhibitor of the Trypanosoma cruzi enolase

Enolase is a glycolytic enzyme that catalyzes the interconversion between 2-phosphoglycerate and phosphoenolpyruvate. In trypanosomatids, enolase was proposed as a key enzyme after in silico and in vivo analysis and it was validated as a protein essential for the survival of the parasite. Therefore, enolase constitutes an interesting enzyme target for the identification of drugs against Chagas disease. In this work, a combined virtual screening strategy was implemented, employing similarity virtual screening, molecular docking, and molecular dynamics. First, two known enolase inhibitors and the enzyme substrates were used as queries for the similarity screening on the Sweetlead database using five different algorithms. Compounds retrieved in the top 10 of at least three search algorithms were selected for further analysis, resulting in six compounds of medical use (etidronate, pamidronate, fosfomycin, acetohydroxamate, triclofos, and aminohydroxybutyrate). Molecular docking simulations and pose re-scoring predicted that binding with acetohydroxamate and triclofos would be weak, while fosfomycin and aminohydroxybutyrate predicted binding is experimentally implausible. Docking poses obtained for etidronate, pamidronate, and PEP were used for molecular dynamics calculations to describe their mode of binding. From the obtained results, we propose etidronate as a potential TcENO inhibitor and describe molecular motifs to be taken into account in the repurposing or design of drugs targeting this enzyme active site.

Identification of a Novel Scaffold for Inhibition of Dipeptidyl Peptidase-4

Dipeptidyl peptidase-4 (DPP-4) is considered a major drug target for type 2 diabetes mellitus (T2DM). In addition to T2DM, a regulatory role of DPP-4 was also found in cardiovascular diseases. Existing DPP-4 inhibitors have been reported to have several adverse effects. In this study, a computer-aided drug design approach and its use to detect a novel class of inhibitor for DPP-4 are reported. Through structure and pharmacophore-based screening, we identified 13 hit compounds from an ∼4-million-compound library. Physical interactions of these hits with DPP-4 were studied using docking and explicit solvent molecular dynamics (MD) simulations. Later, MMPBSA binding energy was calculated for the ligand/protein simulation trajectories to determine the stability of compounds in the binding cavity. These compounds have a novel scaffold and exhibited a stable binding mode. "Best-in-screen" compounds (or their closest available analogs) were resourced and their inhibition of DPP-4 activity was experimentally validated using an in vitro enzyme activity assay in the presence of 100 and 10 μM compounds. These assays identified a compound with a spirochromanone center with 53% inhibition activity at a 100 μM concentration. A further five spirochromanone compounds were synthesized and examined in silico and in vitro; again, one compound showed 53% inhibitory activity action at 100 μM. Overall, this study identified two novel "spirochromanone" compounds that lowered DPP-4 activity by more than ∼50% at 100 μM. This study also showed the impact of fast in silico drug design techniques utilizing virtual screening and MD to identify novel scaffolds to bind and inhibit DPP-4. Spirochromanone motif identified here may be used to design molecules to achieve drug-like inhibitory action against DPP-4.

Comparative Analysis of Electrostatic Models for Ligand Docking

The precise modeling of molecular interactions remains an important goal among molecular modeling techniques. Some of the challenges in the field include the precise definition of a Hamiltonian for biomolecular systems, together with precise parameters derived from Molecular Mechanics Force Fields, for example. The problem is even more challenging when interaction energies from different species are computed, such as the interaction energy involving a ligand and a protein, given that small differences must be computed from large energies. Here we evaluated the effects of the electrostatic model for ligand binding energy evaluation in the context of ligand docking. For this purpose, a classical Coulomb potential with distance-dependent dielectrics was compared with a Poisson-Boltzmann (PB) model for electrostatic potential computation, based on DelPhi calculations. We found that, although the electrostatic energies were highly correlated for the Coulomb and PB models, the ligand pose and the enrichment of actual ligands against decoy compounds, were improved when binding energies were computed using PB as compared to the Coulomb model. We observed that the electrostatic energies computed with the Coulomb model were, on average, ten times larger than the energies computed with the PB model, suggesting a strong overestimation of the polar interactions in the Coulomb model. We also found that a slightly smoothed Lennard-Jones potential combined with the PB model resulted in a good compromise between ligand sampling and energetic scoring.

A Diverse Benchmark Based on 3D Matched Molecular Pairs for Validating Scoring Functions

The prediction of protein-ligand interactions and their corresponding binding free energy is a challenging task in structure-based drug design and related applications. Docking and scoring is broadly used to propose the binding mode and underlying interactions as well as to provide a measure for ligand affinity or differentiate between active and inactive ligands. Various studies have revealed that most docking software packages reliably predict the binding mode, although scoring remains a challenge. Here, a diverse benchmark data set of 99 matched molecular pairs (3D-MMPs) with experimentally determined X-ray structures and corresponding binding affinities is introduced. This data set was used to study the predictive power of 13 commonly used scoring functions to demonstrate the applicability of the 3D-MMP data set as a valuable tool for benchmarking scoring functions.

Applying Pose Clustering and MD Simulations To Eliminate False Positives in Molecular Docking

In this work, we developed a computational protocol that employs multiple molecular docking experiments, followed by pose clustering, molecular dynamic simulations (10 ns), and energy rescoring to produce reliable 3D models of antibody-carbohydrate complexes. The protocol was applied to 10 antibody-carbohydrate co-complexes and three unliganded (apo) antibodies. Pose clustering significantly reduced the number of potential poses. For each system, 15 or fewer clusters out of 100 initial poses were generated and chosen for further analysis. Molecular dynamics (MD) simulations allowed the docked poses to either converge or disperse, and rescoring increased the likelihood that the best-ranked pose was an acceptable pose. This approach is amenable to automation and can be a valuable aid in determining the structure of antibody-carbohydrate complexes provided there is no major side chain rearrangement or backbone conformational change in the H3 loop of the CDR regions. Further, the basic protocol of docking a small ligand to a known binding site, clustering the results, and performing MD with a suitable force field is applicable to any protein ligand system.

Enhance the performance of current scoring functions with the aid of 3D protein-ligand interaction fingerprints

BACKGROUND

In structure-based drug design, binding affinity prediction remains as a challenging goal for current scoring functions. Development of target-biased scoring functions provides a new possibility for tackling this problem, but this approach is also associated with certain technical difficulties. We previously reported the Knowledge-Guided Scoring (KGS) method as an alternative approach (BMC Bioinformatics, 2010, 11, 193-208). The key idea is to compute the binding affinity of a given protein-ligand complex based on the known binding data of an appropriate reference complex, so the error in binding affinity prediction can be reduced effectively.

RESULTS

In this study, we have developed an upgraded version, i.e. KGS2, by employing 3D protein-ligand interaction fingerprints in reference selection. KGS2 was evaluated in combination with four scoring functions (X-Score, ChemPLP, ASP, and GoldScore) on five drug targets (HIV-1 protease, carbonic anhydrase 2, beta-secretase 1, beta-trypsin, and checkpoint kinase 1). In the in situ scoring test, considerable improvements were observed in most cases after application of KGS2. Besides, the performance of KGS2 was always better than KGS in all cases. In the more challenging molecular docking test, application of KGS2 also led to improved structure-activity relationship in some cases.

CONCLUSIONS

KGS2 can be applied as a convenient "add-on" to current scoring functions without the need to re-engineer them, and its application is not limited to certain target proteins as customized scoring functions. As an interpolation method, its accuracy in principle can be improved further with the increasing knowledge of protein-ligand complex structures and binding affinity data. We expect that KGS2 will become a practical tool for enhancing the performance of current scoring functions in binding affinity prediction. The KGS2 software is available upon contacting the authors.

Molecular Docking: From Lock and Key to Combination Lock

Accurate modeling of protein ligand binding is an important step in structure-based drug design, is a useful starting point for finding new lead compounds or drug candidates. The 'Lock and Key' concept of protein-ligand binding has dominated descriptions of these interactions, and has been effectively translated to computational molecular docking approaches. In turn, molecular docking can reveal key elements in protein-ligand interactions-thereby enabling design of potent small molecule inhibitors directed against specific targets. However, accurate predictions of binding pose and energetic remain challenging problems. The last decade has witnessed more sophisticated molecular docking approaches to modeling protein-ligand binding and energetics. However, the complexities that confront accurate modeling of binding phenomena remain formidable. Subtle recognition and discrimination patterns governed by three-dimensional features and microenvironments of the active site play vital roles in consolidating the key intermolecular interactions that mediates ligand binding. Herein, we briefly review contemporary approaches and suggest that future approaches treat protein-ligand docking problems in the context of a 'combination lock' system.

Docking Methodologies and Recent Advances

Docking, a molecular modelling method, has wide applications in identification and optimization in modern drug discovery. This chapter addresses the recent advances in the docking methodologies like fragment docking, covalent docking, inverse docking, post processing, hybrid techniques, homology modeling etc. and its protocol like searching and scoring functions. Advances in scoring functions for e.g. consensus scoring, quantum mechanics methods, clustering and entropy based methods, fingerprinting, etc. are used to overcome the limitations of the commonly used force-field, empirical and knowledge based scoring functions. It will cover crucial necessities and different algorithms of docking and scoring. Further different aspects like protein flexibility, ligand sampling and flexibility, and the performance of scoring function will be discussed.

Screening Explorer-An Interactive Tool for the Analysis of Screening Results

Screening Explorer is a web-based application that allows for an intuitive evaluation of the results of screening experiments using complementary metrics in the field. The usual evaluation of screening results implies the separate generation and apprehension of the ROC, predictiveness, and enrichment curves and their global metrics. Similarly, partial metrics need to be calculated repeatedly for different fractions of a data set and there exists no handy tool that allows reading partial metrics simultaneously on different charts. For a deeper understanding of the results of screening experiments, we rendered their analysis straightforward by linking these metrics interactively in an interactive usable web-based application. We also implemented simple consensus scoring methods based on scores normalization, standardization (z-scores), and compounds ranking to evaluate the enrichments that can be expected through methods combination. Two demonstration data sets allow the users to easily apprehend the functions of this tool that can be applied to the analysis of virtual and experimental screening results. Screening Explorer is freely accessible at http://stats.drugdesign.fr .

Comprehensive evaluation of ten docking programs on a diverse set of protein-ligand complexes: the prediction accuracy of sampling power and scoring power

As one of the most popular computational approaches in modern structure-based drug design, molecular docking can be used not only to identify the correct conformation of a ligand within the target binding pocket but also to estimate the strength of the interaction between a target and a ligand. Nowadays, as a variety of docking programs are available for the scientific community, a comprehensive understanding of the advantages and limitations of each docking program is fundamentally important to conduct more reasonable docking studies and docking-based virtual screening. In the present study, based on an extensive dataset of 2002 protein-ligand complexes from the PDBbind database (version 2014), the performance of ten docking programs, including five commercial programs (LigandFit, Glide, GOLD, MOE Dock, and Surflex-Dock) and five academic programs (AutoDock, AutoDock Vina, LeDock, rDock, and UCSF DOCK), was systematically evaluated by examining the accuracies of binding pose prediction (sampling power) and binding affinity estimation (scoring power). Our results showed that GOLD and LeDock had the best sampling power (GOLD: 59.8% accuracy for the top scored poses; LeDock: 80.8% accuracy for the best poses) and AutoDock Vina had the best scoring power (rp/rs of 0.564/0.580 and 0.569/0.584 for the top scored poses and best poses), suggesting that the commercial programs did not show the expected better performance than the academic ones. Overall, the ligand binding poses could be identified in most cases by the evaluated docking programs but the ranks of the binding affinities for the entire dataset could not be well predicted by most docking programs. However, for some types of protein families, relatively high linear correlations between docking scores and experimental binding affinities could be achieved. To our knowledge, this study has been the most extensive evaluation of popular molecular docking programs in the last five years. It is expected that our work can offer useful information for the successful application of these docking tools to different requirements and targets.

A Real-World Perspective on Molecular Design

We present a series of small molecule drug discovery case studies where computational methods were prospectively employed to impact Roche research projects, with the aim of highlighting those methods that provide real added value. Our brief accounts encompass a broad range of methods and techniques applied to a variety of enzymes and receptors. Most of these are based on judicious application of knowledge about molecular conformations and interactions: filling of lipophilic pockets to gain affinity or selectivity, addition of polar substituents, scaffold hopping, transfer of SAR, conformation analysis, and molecular overlays. A case study of sequence-driven focused screening is presented to illustrate how appropriate preprocessing of information enables effective exploitation of prior knowledge. We conclude that qualitative statements enabling chemists to focus on promising regions of chemical space are often more impactful than quantitative prediction.

Vina-Carb: Improving Glycosidic Angles during Carbohydrate Docking

Molecular docking programs are primarily designed to align rigid, drug-like fragments into the binding sites of macromolecules and frequently display poor performance when applied to flexible carbohydrate molecules. A critical source of flexibility within an oligosaccharide is the glycosidic linkages. Recently, Carbohydrate Intrinsic (CHI) energy functions were reported that attempt to quantify the glycosidic torsion angle preferences. In the present work, the CHI-energy functions have been incorporated into the AutoDock Vina (ADV) scoring function, subsequently termed Vina-Carb (VC). Two user-adjustable parameters have been introduced, namely, a CHI- energy weight term (chi_coeff) that affects the magnitude of the CHI-energy penalty and a CHI-cutoff term (chi_cutoff) that negates CHI-energy penalties below a specified value. A data set consisting of 101 protein-carbohydrate complexes and 29 apoprotein structures was used in the development and testing of VC, including antibodies, lectins, and carbohydrate binding modules. Accounting for the intramolecular energies of the glycosidic linkages in the oligosaccharides during docking led VC to produce acceptable structures within the top five ranked poses in 74% of the systems tested, compared to a success rate of 55% for ADV. An enzyme system was employed in order to illustrate the potential application of VC to proteins that may distort glycosidic linkages of carbohydrate ligands upon binding. VC represents a significant step toward accurately predicting the structures of protein-carbohydrate complexes. Furthermore, the described approach is conceptually applicable to any class of ligands that populate well-defined conformational states.

Insight into human protease activated receptor-1 as anticancer target by molecular modelling

Protease-activated receptor 1 (PAR1) has been established as a promising target in many diseases, including various cancers. Strong evidence also suggests its role in metastasis. It is proved experimentally that PAR1 can induce numerous cell phenotypes, i.e. proliferation and differentiation. A strong link between PAR1 gene overexpression and high levels of ß-catenin was suggested by a study of the PAR1-Gα(13)-DVL axis in ß-catenin stabilization in cancers. An in vitro study was carried out to analyze PAR1 expression by flow cytometry on CD38+138+ plasma cells obtained from patients either at diagnosis (n: 46) (newly diagnosed multiple myeloma (NDMM)) or at relapse (n: 45) (relapsed/refractory multiple myeloma (RRMM)) and compared with the controls. Our previously synthesized benzoxazole (XT2B) and benzamide (XT5) derivatives were tested with in vitro 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assays, which revealed significant inhibitory activity on PAR1. We provide docking studies using Autodock Vina of these newly tested compounds to compare with the known PAR1 inhibitors in order to examine the binding mechanisms. In addition, the docking results are validated using HYDE binding assessment and a neural network (NN) scoring function.

Charting a Path to Success in Virtual Screening

Docking is commonly applied to drug design efforts, especially high-throughput virtual screenings of small molecules, to identify new compounds that bind to a given target. Despite great advances and successful applications in recent years, a number of issues remain unsolved. Most of the challenges and problems faced when running docking experiments are independent of the specific software used, and can be ascribed to either improper input preparation or to the simplified approaches applied to achieve high-throughput speed. Being aware of approximations and limitations of such methods is essential to prevent errors, deal with misleading results, and increase the success rate of virtual screening campaigns. In this review, best practices and most common issues of docking and virtual screening will be discussed, covering the journey from the design of the virtual experiment to the hit identification.

Comparative assessment of scoring functions on an updated benchmark: 1. Compilation of the test set

Scoring functions are often applied in combination with molecular docking methods to predict ligand binding poses and ligand binding affinities or to identify active compounds through virtual screening. An objective benchmark for assessing the performance of current scoring functions is expected to provide practical guidance for the users to make smart choices among available methods. It can also elucidate the common weakness in current methods for future improvements. The primary goal of our comparative assessment of scoring functions (CASF) project is to provide a high-standard, publicly accessible benchmark of this type. Our latest study, i.e., CASF-2013, evaluated 20 popular scoring functions on an updated set of protein-ligand complexes. This data set was selected out of 8302 protein-ligand complexes recorded in the PDBbind database (version 2013) through a fairly complicated process. Sample selection was made by considering the quality of complex structures as well as binding data. Finally, qualified complexes were clustered by 90% similarity in protein sequences. Three representative complexes were chosen from each cluster to control sample redundancy. The final outcome, namely, the PDBbind core set (version 2013), consists of 195 protein-ligand complexes in 65 clusters with binding constants spanning nearly 10 orders of magnitude. In this data set, 82% of the ligand molecules are "druglike" and 78% of the protein molecules are validated or potential drug targets. Correlation between binding constants and several key properties of ligands are discussed. Methods and results of the scoring function evaluation will be described in a companion work in this issue (doi: 10.1021/ci500081m ).

Comparative assessment of scoring functions on an updated benchmark: 2. Evaluation methods and general results

Our comparative assessment of scoring functions (CASF) benchmark is created to provide an objective evaluation of current scoring functions. The key idea of CASF is to compare the general performance of scoring functions on a diverse set of protein-ligand complexes. In order to avoid testing scoring functions in the context of molecular docking, the scoring process is separated from the docking (or sampling) process by using ensembles of ligand binding poses that are generated in prior. Here, we describe the technical methods and evaluation results of the latest CASF-2013 study. The PDBbind core set (version 2013) was employed as the primary test set in this study, which consists of 195 protein-ligand complexes with high-quality three-dimensional structures and reliable binding constants. A panel of 20 scoring functions, most of which are implemented in main-stream commercial software, were evaluated in terms of "scoring power" (binding affinity prediction), "ranking power" (relative ranking prediction), "docking power" (binding pose prediction), and "screening power" (discrimination of true binders from random molecules). Our results reveal that the performance of these scoring functions is generally more promising in the docking/screening power tests than in the scoring/ranking power tests. Top-ranked scoring functions in the scoring power test, such as X-Score(HM), ChemScore@SYBYL, ChemPLP@GOLD, and PLP@DS, are also top-ranked in the ranking power test. Top-ranked scoring functions in the docking power test, such as ChemPLP@GOLD, Chemscore@GOLD, GlidScore-SP, LigScore@DS, and PLP@DS, are also top-ranked in the screening power test. Our results obtained on the entire test set and its subsets suggest that the real challenge in protein-ligand binding affinity prediction lies in polar interactions and associated desolvation effect. Nonadditive features observed among high-affinity protein-ligand complexes also need attention.

What has computer-aided molecular design ever done for drug discovery?

This article assesses the contribution of computer-aided molecular design (CAMD) to the field of drug discovery. Several examples of ligand- and structure-based drug design are used to demonstrate the role of CAMD in the discovery of marketed drug compounds. Although CAMD is now an integral part of many drug discovery projects, there are significant challenges still facing its practitioners, particularly the prediction of binding affinity.

RANK-ORDERING THE BINDING AFFINITY FOR FKBP12 AND H1N1 NEURAMINIDASE INHIBITORS IN THE COMBINATION OF A PROTEIN MODEL WITH DENSITY FUNCTIONAL THEORY

The quantum mechanical interaction energies between FKBP12 as well as H1N1 neuraminidase and their inhibitors were directly calculated with an efficient density functional theory by mimicking the whole protein with a protein model composed of the amino acids surrounding the ligands. It was found that the calculated quantum mechanical interaction energies correlate well with the experimental binding free energies with the correlation coefficients of 0.88, 0.86, and the standard deviation of 0.93 and 1.00 kcal/mol, respectively. To compare with force field approach, the binding free energies with the correlation coefficient R = 0.80 and 0.47 were estimated by AutoDock 4.0 programs. It was indicated that the quantum interaction energy shows a better performance in rank-ordering the binding affinity between FKBP12 and H1N1 neuraminidase inhibitors than those of AutoDock 4.0 program. In combination protein model with density functional theory, the estimated quantum interaction energy could be a good predictor or scoring function in structure-based computer-aided drug design. Finally, five new FKBP12 inhibitors were designed based on calculated quantum mechanical interaction energy. In particular, the theoretical K i value of one compound is as low as 0.05 nM, nearly 8-fold more active than FK506.

NNScore: a neural-network-based scoring function for the characterization of protein-ligand complexes

As high-throughput biochemical screens are both expensive and labor intensive, researchers in academia and industry are turning increasingly to virtual-screening methodologies. Virtual screening relies on scoring functions to quickly assess ligand potency. Although useful for in silico ligand identification, these scoring functions generally give many false positives and negatives; indeed, a properly trained human being can often assess ligand potency by visual inspection with greater accuracy. Given the success of the human mind at protein−ligand complex characterization, we present here a scoring function based on a neural network, a computational model that attempts to simulate, albeit inadequately, the microscopic organization of the brain. Computer-aided drug design depends on fast and accurate scoring functions to aid in the identification of small-molecule ligands. The scoring function presented here, used either on its own or in conjunction with other more traditional functions, could prove useful in future drug-discovery efforts.

Test MM-PB/SA on true conformational ensembles of protein-ligand complexes

The molecular mechanics Poisson-Boltzmann surface area (MM-PB/SA) method has been popular for computing protein-ligand binding free energies in recent years. All previous evaluations of the MM-PB/SA method are based upon computer-generated conformational ensembles, which may be affected by the defective computational methods used for preparing these conformational ensembles. In an attempt to reach more convincing conclusions, we have evaluated the MM-PB/SA method on a set of 24 diverse protein-ligand complexes, each of which has a set of conformations derived from NMR spectroscopy. Our results indicate that both MM-PB/SA and molecular mechanics generalized Born surface area (MM-GB/SA) are able to produce a modest correlation between their results and the experimentally measured binding free energies on our test set. In particular, both MM-PB/SA and MM-GB/SA produced better results by using a representative structure (R = 0.72-0.79) rather than averaging over the conformational ensemble of each given complex (R = 0.61-0.74). A head-to-head comparison with four selected scoring functions (X-Score, PLP, ChemScore, and DrugScore) on the same test set reveals that MM-PB/SA and MM-GB/SA results are marginally better than those produced by scoring funcitons, supporting the value of the MM-PB/SA method. Nevertheless, scoring functions are still more cost-effective options, especially for high-throughput tasks.

Some sulfonamide drugs inhibit ATPase activity of heat shock protein 90: investigation by docking simulation and experimental validation

Eight selected sulfonamide drugs were investigated as inhibitors of heat shock protein 90 (Hsp90). The investigation included simulated docking experiments to fit the selected compounds within the binding pocket of Hsp90. The selected molecules were found to readily fit within the ATP-binding pocket of Hsp90 in low-energy poses. The sulfonamides torsemide, sulfathiazole, and sulfadiazine were found to inhibit the ATPase activity of Hsp90 with IC(50) values of 1.0, 2.6, and 1.5 μM, respectively. Our results suggest that these well-established sulfonamides can be good leads for subsequent optimization into potent Hsp90 inhibitors.

Scoring functions and their evaluation methods for protein-ligand docking: recent advances and future directions

The scoring function is one of the most important components in structure-based drug design. Despite considerable success, accurate and rapid prediction of protein-ligand interactions is still a challenge in molecular docking. In this perspective, we have reviewed three basic types of scoring functions (force-field, empirical, and knowledge-based) and the consensus scoring technique that are used for protein-ligand docking. The commonly-used assessment criteria and publicly available protein-ligand databases for performance evaluation of the scoring functions have also been presented and discussed. We end with a discussion of the challenges faced by existing scoring functions and possible future directions for developing improved scoring functions.

Advances and challenges in protein-ligand docking

Molecular docking is a widely-used computational tool for the study of molecular recognition, which aims to predict the binding mode and binding affinity of a complex formed by two or more constituent molecules with known structures. An important type of molecular docking is protein-ligand docking because of its therapeutic applications in modern structure-based drug design. Here, we review the recent advances of protein flexibility, ligand sampling, and scoring functions—the three important aspects in protein-ligand docking. Challenges and possible future directions are discussed in the Conclusion.

Maximum unbiased validation (MUV) data sets for virtual screening based on PubChem bioactivity data

Refined nearest neighbor analysis was recently introduced for the analysis of virtual screening benchmark data sets. It constitutes a technique from the field of spatial statistics and provides a mathematical framework for the nonparametric analysis of mapped point patterns. Here, refined nearest neighbor analysis is used to design benchmark data sets for virtual screening based on PubChem bioactivity data. A workflow is devised that purges data sets of compounds active against pharmaceutically relevant targets from unselective hits. Topological optimization using experimental design strategies monitored by refined nearest neighbor analysis functions is applied to generate corresponding data sets of actives and decoys that are unbiased with regard to analogue bias and artificial enrichment. These data sets provide a tool for Maximum Unbiased Validation (MUV) of virtual screening methods. The data sets and a software package implementing the MUV design workflow are freely available at http://www.pharmchem.tu-bs.de/lehre/baumann/MUV.html.

Hybrid structure-based virtual screening protocol for the identification of novel BACE1 inhibitors

BACE1, also called beta-secretase or memapsin 2, is an extensively studied aspartic protease, involved in etiopathogenesis and progression of Alzheimer's disease (AD). We report herein a modified structure-based virtual screening protocol that augments the lead identification process against BACE1 during virtual screening endeavors. A hybrid structure-based virtual screening protocol that incorporates elements from both ligand-based and structure-based techniques was used for the identification of prospective small molecule inhibitors. Virtual screening, using an active-site-derived pharmacophore, followed by ROCS (rapid overlay of chemical structures)-based GOLD (genetic optimization in ligand docking) docking was used to identify a library of focused candidates. The efficacy of the ROCS-based GOLD docking method together with our customized weighted consensus scoring function was evaluated against conventional docking methods for its ability to discern true positives from a screening library. An in-depth structural analysis of the binding mode of the top-ranking molecules reveals that emulation of the curial interaction patterns deemed necessary for BACE1 inhibition. The results obtained from our validation study ensure the superiority of our docking methodology over conventional docking methods in yielding higher enrichment rates.

Scoring confidence index: statistical evaluation of ligand binding mode predictions

Protein-ligand docking programs can generate a large number of possible binding orientations for each ligand candidate. The challenge is to identify the orientations closest to the native binding mode using a scoring method. Many different scoring functions have been developed for protein-ligand scoring, but their performance on binding mode prediction is often target-dependent. In this study, a statistical approach was employed to provide a confidence measure of scoring performance in finding close to the correct docked ligand orientations. It exploits the fact that the scores provided by an adequately performing scoring function generally improve as the ligand binding modes get closer to the correct native orientation. For such cases, the correlation coefficient of scores versus distances is expected to be highest when the most native-like orientation is used as a reference. This correlation coefficient, called the correlation-based score (CBScore), was used as an indicator of how far the docked pose was from the native orientation. The correlation between the original scores and CBScores as well as the range of CBScores were found to be good measures of scoring performance. They were combined into a single quantity, called the scoring confidence index. High values of the scoring confidence index were indicative of pronounced and relatively smooth binding energy landscapes with easily discernable global minima, resulting in reliable binding mode predictions. Low values of this index reflected rugged energy landscapes making the prediction of the correct binding mode very difficult and often unreliable. The diagnostic ability of the scoring confidence index was tested on a non-redundant set of 50 protein-ligand complexes scored with three commonly employed scoring functions: AffiScore, DrugScore and X-Score. Binding mode predictions were found to be three times more reliable for complexes with scoring confidence indices in the upper half than for cases with values in the lower half of the resulting range of 0-1.6. This new confidence measure of scoring performance is expected to be a valuable tool for virtual screening applications.

Modeling of peroxide activation in artemisinin derivatives by serial docking

Serial docking of artemisinin derivatives. A serial docking study was undertaken with the purpose to improve the understanding of the mechanism of production of biological activity in Artemisinin derivatives. The Heme molecule receptors were primarily chosen to represent the changing binding and oxidation states of this molecule, which are postulated to occur during the activation of the drug, in order to relate these results to the observed biological activity. The results of the docking runs were classified according to similarity to a standard orientation by a combination of automated and "by hand" procedures. One- and two-dimensional QSAR equations were used in an exploratory sense in order to study the influence of the type of receptor. Principal component and partial least squares regression techniques were used in the case of the multivariate (3D-QSAR) descriptors. The results obtained corroborate the postulated mechanism of production of biological activity as well as providing evidence that, at the moment of activation, the electronic structure of the Heme molecule approximates that of the oxygenated Heme, the drug molecule adopts a preferred orientation, and that, in addition to the important positive contribution of the O(1) - Fe interaction, there is as well a significant negative effect on the biological activity by carbons 4 - 6 of the artemisinin ring system.