Ligand-based Model Research Articles

Predicting Off-Target Binding Profiles With Confidence Using Conformal Prediction.

Ligand-based models can be used in drug discovery to obtain an early indication of potential off-target interactions that could be linked to adverse effects. Another application is to combine such models into a panel, allowing to compare and search for compounds with similar profiles. Most contemporary methods and implementations however lack valid measures of confidence in their predictions, and only provide point predictions. We here describe a methodology that uses Conformal Prediction for predicting off-target interactions, with models trained on data from 31 targets in the ExCAPE-DB dataset selected for their utility in broad early hazard assessment. Chemicals were represented by the signature molecular descriptor and support vector machines were used as the underlying machine learning method. By using conformal prediction, the results from predictions come in the form of confidence p-values for each class. The full pre-processing and model training process is openly available as scientific workflows on GitHub, rendering it fully reproducible. We illustrate the usefulness of the developed methodology on a set of compounds extracted from DrugBank. The resulting models are published online and are available via a graphical web interface and an OpenAPI interface for programmatic access.

Self-cleavage of the glmS ribozyme core is controlled by a fragile folding element

Riboswitches modulate gene expression in response to small-molecule ligands. Switching is generally thought to occur via the stabilization of a specific RNA structure conferred by binding the cognate ligand. However, it is unclear whether any such stabilization occurs for riboswitches whose ligands also play functional roles, such as the glmS ribozyme riboswitch, which undergoes self-cleavage using its regulatory ligand, glucosamine 6-phosphate, as a catalytic cofactor. To address this question, it is necessary to determine both the conformational ensemble and its ligand dependence. We used optical tweezers to measure folding dynamics and cleavage rates for the core glmS ribozyme over a range of forces and ligand conditions. We found that the folding of a specific structural element, the P2.2 duplex, controls active-site formation and catalysis. However, the folded state is only weakly stable, regardless of cofactor concentration, supplying a clear exception to the ligand-based stabilization model of riboswitch function.

Evaluating parameters for ligand-based modeling with random forest on sparse data sets

Ligand-based predictive modeling is widely used to generate predictive models aiding decision making in e.g. drug discovery projects. With growing data sets and requirements on low modeling time comes the necessity to analyze data sets efficiently to support rapid and robust modeling. In this study we analyzed four data sets and studied the efficiency of machine learning methods on sparse data structures, utilizing Morgan fingerprints of different radii and hash sizes, and compared with molecular signatures descriptor of different height. We specifically evaluated the effect these parameters had on modeling time, predictive performance, and memory requirements using two implementations of random forest; Scikit-learn as well as FEST. We also compared with a support vector machine implementation. Our results showed that unhashed fingerprints yield significantly better accuracy than hashed fingerprints (p le 0.05), with no pronounced deterioration in modeling time and memory usage. Furthermore, the fast execution and low memory usage of the FEST algorithm suggest that it is a good alternative for large, high dimensional sparse data. Both support vector machines and random forest performed equally well but results indicate that the support vector machine was better at using the extra information from larger values of the Morgan fingerprint’s radius.

Probability-driven 3D pharmacophore mapping of antimycobacterial potential of hybrid molecules combining phenylcarbamoyloxy and N-arylpiperazine fragments

ABSTRACTThe current study examines in silico characterization of the structure-inhibitory potency for a set of phenylcarbamic acid derivatives containing an N-arylpiperazine scaffold, considering the electronic, steric and lipophilic properties. The main objective of the ligand-based modelling was the systematic study of classical comparative molecular field analysis (CoMFA)/comparative molecular surface analysis (CoMSA) performance for the modelling of in vitro efficiency observed for these phenylcarbamates, revealing their inhibitory activities against a virulent Mycobacterium tuberculosis H37Rv strain. We compared the findings of efficiency modelling produced by a standard 3D methodology (CoMFA) and its neural counterparts (CoMSA) regarding multiple training/test subsets and variables used. Moreover, systematic space inspection, splitting values into the analysed training/test subsets, was performed to monitor statistical estimator performance while mapping the probability-driven pharmacophore pattern. Consequently, a ‘pseudo-consensus’ 3D-quantitative structure-activity relationship (3D-QSAR) approach was applied to retrieve an ‘average’ pharmacophore hypothesis by the investigation of the most densely populated training/test subpopulations to specify the potentially important factors contributing to the inhibitory activity of phenylcarbamic acid analogues. In addition, examination of descriptor-based similarity with a principal component analysis (PCA) procedure was employed to visualize noticeable variations in the performance of these molecules with respect to their structure and activity profiles.

Ligand-Mediated Nucleation and Growth of Palladium Metal Nanoparticles.

The size, size distribution and stability of colloidal nanoparticles are greatly affected by the presence of capping ligands. Despite the key contribution of capping ligands during the synthesis reaction, their role in regulating the nucleation and growth rates of colloidal nanoparticles is not well understood. In this work, we demonstrate a mechanistic investigation of the role of trioctylphosphine (TOP) in Pd nanoparticles in different solvents (toluene and pyridine) using in situ SAXS and ligand-based kinetic modeling. Our results under different synthetic conditions reveal the overlap of nucleation and growth of Pd nanoparticles during the reaction, which contradicts the LaMer-type nucleation and growth model. The model accounts for the kinetics of Pd-TOP binding for both, the precursor and the particle surface, which is essential to capture the size evolution as well as the concentration of particles in situ. In addition, we illustrate the predictive power of our ligand-based model through designing the synthetic conditions to obtain nanoparticles with desired sizes. The proposed methodology can be applied to other synthesis systems and therefore serves as an effective strategy for predictive synthesis of colloidal nanoparticles.

Development of Ligand-based Big Data Deep Neural Network Models for Virtual Screening of Large Compound Libraries.

High-performance ligand-based virtual screening (VS) models have been developed using various computational methods, including the deep neural network (DNN) method. There are high expectations for exploration of the advanced capabilities of DNN to improve VS performance, and this capability has been optimally achieved using large data training datasets. However, their ability to screen large compound libraries has not been evaluated. There is a need for developing and evaluating ligand-based large data DNN VS models for large compound libraries. In this study, we developed ligand-based large data DNN VS models for inhibitors of six anticancer targets using 0.5 M training compounds. The developed VS models were evaluated by 10-fold cross-validation, achieving 77.9-97.8 % sensitivity, 99.9-100 % specificity, 0.82-0.98 Matthews correlation coefficient and 0.98-0.99 area under the curve, outperforming random forest models. Moreover, DNN VS models developed by pre-2015 inhibitors identified 50 % of post-2015 inhibitors with a 0.01-0.09 % false positive rate in screening 89 M PubChem compounds, also outperforming previous models. Experimental assays of the selected virtual hits of the EGFR inhibitor model led to reasonable novel structures of EGFR inhibitors. Our results confirmed the usefulness of the large data DNN model as a ligand-based VS tool to screen large compound libraries.

Modeling Kinase Inhibition Using Highly Confident Data Sets.

2016 , 56 , 895 - 905 ; www.chembioinf.ro ) aiming to build and measure the performance of ligand-based kinase inhibitor prediction models. Here we analyzed kinase-inhibitor pairs with multiple activity points extracted from the ChEMBL database and identified the main sources of inconsistency. Our results indicate that lower IC50 values are usually less affected by errors and reflect more accurately the structure-activity relationship of the molecules against the target, ideally for quantitative structure-activity relationship studies. Further, we modeled the activities of 104 kinases using unbiased target-specific activity points. The performance of predictors built on extended connectivity fingerprints (ECFP4) and two-dimensional pharmacophore fingerprints (PFPs) are compared by means of tolerance intervals (TIs) (95%/95%) in virtual screening (VS) and classification tasks using external random ( RandSets) and diversity-based ( DivSets) test sets. We found that the two encodings perform superior to each other on different kinases in VS and that PFP models perform consistently better in classifying actives (higher sensitivity). Next, we combined the two encodings into a single one (PFPECFP) and demonstrated that especially in VS (as indicated by the exponential receiver operating curve enrichment metric (eROCE)), for the vast majority of kinases the model performance increased compared with the individual fingerprint models. These findings are highlighted in the more challenging DivSets compared with RandSets. The current paper explores the boundaries of inhibitor predictors for individual kinases to enhance VS and ultimately aid the discovery of novel compounds with desirable polypharmacology.

Ligand and Structure-based Modeling of Passive Diffusion through the Blood-Brain Barrier.

Blood-brain barrier transport is an important process to be considered in drug candidates. The blood-brain barrier protects the brain from toxicological agents and, therefore, also establishes a restrictive mechanism for the delivery of drugs into the brain. Although there are different and complex mechanisms implicated in drug transport, in this review we focused on the prediction of passive diffusion through the blood-brain barrier. We elaborated on ligand-based and structure-based models that have been described to predict the blood-brain barrier permeability. Multiple 2D and 3D QSPR/QSAR models and integrative approaches have been published to establish quantitative and qualitative relationships with the blood-brain barrier permeability. We explained different types of descriptors that correlate with passive diffusion along with data analysis methods. Moreover, we discussed the applicability of other types of molecular structure-based simulations, such as molecular dynamics, and their implications in the prediction of passive diffusion. Challenges and limitations of experimental measurements of permeability and in silico predictive methods were also described. Improvements in the prediction of blood-brain barrier permeability from different types of in silico models are crucial to optimize the process of Central Nervous System drug discovery and development.

Exploration of structural and physicochemical properties of small molecules to inhibit NMDA functionality

The N-methyl-D-aspartate (NMDA) is the family of glutamate receptor, which is involved in controlling synaptic plasticity and memory function; but overactivation of this receptor results to excess intracellular calcium formation, triggers neuronal injury and also involves in several pathologies. Both ligand- and structure-based quantitative structure-activity relationship (QSAR), pharmacophore, docking and simulation studies have been performed on a set of structurally diverse inhibitors to explore prime molecular structural features involve for specific binding to NMDA, and vis-a-vis inhibiting enzyme activity. 3D QSAR studies, comparative molecular field analysis (CoMFA) and comparative molecular similarity indices analysis (CoMSIA) models showed the importance of steric, electrostatic and hydrophobic features; while hydrogen bond acceptor and hydrophobic features are depicted as important pharmacophore features of the molecule. Molecular docking and simulation studies corroborated the consequence of the features obtained from ligand-based Bayesian model (AUROCcv = 0.878); 3D QSAR CoMFA (R2 = 0.895, se = 0.513, Q2 = 0.602, R2pred = 0.673); CoMSIA (R2 = 0.877, se = 0.555, Q2 = 0.615, R2pred = 0.727); hologram QSAR (Q2 = 0.812, R2 = 0.941, R2pred = 0.772), and pharmacophore models (Q2 = 0.926, R2 = 0.927, R2pred = 0.621). Presence of aromatic ring, hetero and halogen atoms along with alkyl group of molecular scaffold shows their importance for binding affinity to NMDA receptor. Stability of the complex is adjudged by both docking and simulation studies.

Targeting Cysteine Proteases from Plasmodium falciparum: A General Overview, Rational Drug Design and Computational Approaches for Drug Discovery.

The Plasmodium falciparum cysteine proteases, also known as falcipains, are involved in different erythrocytic cycle processes of the malaria parasite, e.g. hydrolysis of host haemoglobin, erythrocyte invasion, and erythrocyte rupture. With the biochemical characterization of four falcipains so far, FP-2 (falcipain-2) and FP-3 (falcipain-3), members of the papain-like CAC1 family, are essential haemoglobinases. They could therefore be referred to as potential anti-malarial drug targets in the search for novel therapies, which could ease the burden caused by the increasing resistance to current antimalarial drugs. This review provides a summary of the most important results, highlighting the drug design approaches essential for the understanding of the mechanism of inhibition and discovery of inhibitors against cysteine proteases from P. falciparum. Rational and computer-aided drug discovery approaches for the design of promising falcipain inhibitors are described herein, with a focus on a variety of structure-based and ligand-based modeling approaches. Moreover, the key features of ligand recognition against these targets are emphasized. This review would be of interest to scientists engaged in the development of drug design strategies to target the cysteine proteases, FP-2 and FP-3.

Efficient iterative virtual screening with Apache Spark and conformal prediction

BackgroundDocking and scoring large libraries of ligands against target proteins forms the basis of structure-based virtual screening. The problem is trivially parallelizable, and calculations are generally carried out on computer clusters or on large workstations in a brute force manner, by docking and scoring all available ligands.ContributionIn this study we propose a strategy that is based on iteratively docking a set of ligands to form a training set, training a ligand-based model on this set, and predicting the remainder of the ligands to exclude those predicted as ‘low-scoring’ ligands. Then, another set of ligands are docked, the model is retrained and the process is repeated until a certain model efficiency level is reached. Thereafter, the remaining ligands are docked or excluded based on this model. We use SVM and conformal prediction to deliver valid prediction intervals for ranking the predicted ligands, and Apache Spark to parallelize both the docking and the modeling.ResultsWe show on 4 different targets that conformal prediction based virtual screening (CPVS) is able to reduce the number of docked molecules by 62.61% while retaining an accuracy for the top 30 hits of 94% on average and a speedup of 3.7. The implementation is available as open source via GitHub (https://github.com/laeeq80/spark-cpvs) and can be run on high-performance computers as well as on cloud resources.

Ligand-based modeling of Akt3 lead to potent dual Akt1/Akt3 inhibitor

Akt1 and Akt3 are important serine/threonine-specific protein kinases involved in G2 phase required by cancer cells to maintain cell cycle and to prevent cell death. Accordingly, inhibitors of these kinases should have potent anti-cancer properties. This prompted us to use pharmacophore/QSAR modeling to identify optimal binding models and physicochemical descriptors that explain bioactivity variation within a set of 74 diverse Akt3 inhibitors. Two successful orthogonal pharmacophores were identified and further validated using receiver operating characteristic (ROC) curve analyses. The pharmacophoric models and associated QSAR equation were applied to screen the national cancer institute (NCI) list of compounds for new Akt3 inhibitors. Six hits showed significant experimental anti-Akt3 IC50 values, out of which one compound exhibited dual low micromolar anti-Akt1 and anti-Akt3 inhibitory profiles.

Toward the identification of a reliable 3D-QSAR model for the protein tyrosine phosphatase 1B inhibitors

Protein tyrosine phosphatase 1B (PTP1B) is an intracellular non-receptor phosphatase that is implicated in signal transduction of insulin and leptin pathways, thus PTP1B is considered as potential target for treating type II diabetes and obesity. The present article is an attempt to formulate the three-dimensional quantitative structure-activity relationship (3D-QSAR) modeling of a series of compounds possessing PTP1B inhibitory activities using comparative molecular field analysis (CoMFA) and comparative molecular similarity indices analysis (CoMSIA) techniques. The optimum template ligand-based models are statistically significant with great CoMFA (R2cv = 0.600, R2pred = 0.6760) and CoMSIA (R2cv = 0.624, R2pred = 0.8068) values. Molecular docking was employed to elucidate the inhibitory mechanisms of this series of compounds against PTP1B. In addition, the CoMFA and CoMSIA field contour maps agree well with the structural characteristics of the binding pocket of PTP1B active site. The knowledge of structure-activity relationship and ligand-receptor interactions from 3D-QSAR model and molecular docking will be useful for better understanding the mechanism of ligand-receptor interaction and facilitating development of novel compounds as potent PTP1B inhibitors.

In silico ligand-based modeling of hBACE-1 inhibitors.

Alzheimer's disease is a chronic neurodegenerative disease affecting more than 30 million people worldwide. Development of small molecule inhibitors of human β-secretase 1 (hBACE-1) is being the focus of pharmaceutical industry for the past 15-20years. Here, we successfully applied multiple ligand-based in silico modeling techniques to understand the inhibitory activities of a diverse set of small molecule hBACE-1 inhibitors reported in the scientific literature. Strikingly, the use of only a small subset of 230 (13%) molecules allowed us to develop quality models that performed reasonably well on the validation set of 1,476 (87%) inhibitors. Varying the descriptor sets and the complexity of the modeling techniques resulted in only minor improvements to the model's performance. The current results demonstrate that predictive models can be built by choosing appropriate modeling techniques in spite of using small datasets consisting of diverse chemical classes, a scenario typical in triaging of high-throughput screening results to identify false negatives. We hope that these encouraging results will help the community to develop more predictive models that would support research efforts for the debilitating Alzheimer's disease. Additionally, the integrated diversity of the techniques employed will stimulate scientists in the field to use in silico statistical modeling techniques like these to derive better models to help advance the drug discovery projects faster.

Utilizing random Forest QSAR models with optimized parameters for target identification and its application to target-fishing server

BackgroundThe identification of target molecules is important for understanding the mechanism of “target deconvolution” in phenotypic screening and “polypharmacology” of drugs. Because conventional methods of identifying targets require time and cost, in-silico target identification has been considered an alternative solution. One of the well-known in-silico methods of identifying targets involves structure activity relationships (SARs). SARs have advantages such as low computational cost and high feasibility; however, the data dependency in the SAR approach causes imbalance of active data and ambiguity of inactive data throughout targets.ResultsWe developed a ligand-based virtual screening model comprising 1121 target SAR models built using a random forest algorithm. The performance of each target model was tested by employing the ROC curve and the mean score using an internal five-fold cross validation. Moreover, recall rates for top-k targets were calculated to assess the performance of target ranking. A benchmark model using an optimized sampling method and parameters was examined via external validation set. The result shows recall rates of 67.6% and 73.9% for top-11 (1% of the total targets) and top-33, respectively. We provide a website for users to search the top-k targets for query ligands available publicly at http://rfqsar.kaist.ac.kr.ConclusionsThe target models that we built can be used for both predicting the activity of ligands toward each target and ranking candidate targets for a query ligand using a unified scoring scheme. The scores are additionally fitted to the probability so that users can estimate how likely a ligand–target interaction is active. The user interface of our web site is user friendly and intuitive, offering useful information and cross references.

Selectivity and ligand-based molecular modeling of an odorant-binding protein from the leaf beetle Ambrostoma quadriimpressum (Coleoptera: Chrysomelidae) in relation to habitat-related volatiles

In this study, the most abundant and antenna-specific odorant-binding protein (OBP) of the elm pest A. quadriimpressum, AquaOBP4, was expressed and purified. The selectivity of AquaOBP4 was investigated by screening against a panel of 40 habitat-relevant compounds. Based on the obtained results, a homologous model of AquaOBP4 was established. This model indicated that AquaOBP4 is highly homologous to DmelOBP LUSH and includes two main binding sites. A docking analysis showed that four of five active ligands bound at Site 1, whereas the other ligand was situated at Site 2. Furthermore, new ligands were docked in the model, and the results of fluorescence-based binding assays of these compounds were highly consistent with the binding conformation and binding affinity predicted by our model. Additionally, three binding odorants derived from elm leaves elicited a strong electroantennogram response and exerted a significant attractive effect on adult A. quadriimpressum. All of the results showed that AquaOBP4 is likely linked to the foraging behavior of A. quadriimpressum. This study provides a new reliable tool for future large-scale compound screenings and revealed several functional chemicals that might aid the development of a better pest management approach for A. quadriimpressum.

Ligand-based computational modelling of platelet-derived growth factor beta receptor leading to new angiogenesis inhibitory leads

Platelet derived growth factor beta receptor (PDGFR- β) plays an important role in angiogenesis. PDGFR-β expression is correlated with increased vascularity and maturation of blood vessels in cancer. Pharmacophore modeling and quantitative structure-activity relationship (QSAR) analysis were combined to explore the structural requirements for ligand-PDGFR-β recognition using 107 known PDGFR-β inhibitors. Genetic function algorithm (GFA) coupled to k nearest neighbor (kNN) and multiple linear regression (MLR) analysis were employed to generate predictive QSAR models based on optimal combinations of pharmacophores and physicochemical descriptors. The successful pharmacophores were complemented with exclusion spheres to optimize their receiver operating characteristic curve (ROC) profiles. The QSAR models and their associated pharmacophore hypotheses were validated by identification and experimental evaluation of new angiogenesis inhibitory leads retrieved from the National Cancer Institute (NCI) structural database. Two hits illustrated low micromolar IC50 values in two distinct anti-angiogenesis bioassays.

Studies of the benzopyran class of selective COX-2 inhibitors using 3D-QSAR and molecular docking.

The Gaussian-based 3D-QSAR studies for 58 selective COX-2 (cyclooxygenase-2) inhibitors belonging to benzopyran chemical class were performed. Partial least squares analysis produced statistically significant model with (R training 2 =0.866) and predictability (Q training 2 =0.66, Q test 2 =0.846). The 3D-QSAR model includes steric, electrostatic, hydrophobic, and hydrogen bond acceptor field indicators, whereas the potential field contributions indicate that the steric and hydrophobic features of the molecules play an important role in governing their biological activity. A molecular docking simulation and protein-ligand interaction pattern analysis reveal the importance of Tyr-361 and Ser-516 of the COX-2 active site for X-ray crystal structures and this class of molecules. Thus the combined approach of ligand-based and structure-based models provided an improved understanding in the interaction between benzopyran chemical class and COX-2 inhibition, which will guide the future identification of more potent anti-inflammatory drugs.

Molecular properties associated with transporter-mediated drug disposition.

Membrane transporters play a key role in the absorption, distribution, clearance, elimination, and transport of drugs. Understanding the drug properties and structure activity relationships (SAR) for affinity to membrane transporters is critical to optimize clearance and pharmacokinetics during drug design. To facilitate the early identification of clearance mechanism, a framework named the extended clearance classification system (ECCS) was recently introduced. Using in vitro and physicochemical properties that are readily available in early drug discovery, ECCS has been successfully applied to identify major clearance mechanism and to implicate the role of membrane transporters in determining pharmacokinetics. While the crystal structures for most of the drug transporters are currently not available, ligand-based modeling approaches that use information obtained from the structure and molecular properties of the ligands have been applied to associate the drug-related properties and transporter-mediated disposition. The approach allows prospective prediction of transporter both substrate and/or inhibitor affinity and build quantitative structure-activity relationship (QSAR) to enable early optimization of pharmacokinetics, tissue distribution and drug-drug interaction risk. Drug design applications can be further improved through uncovering transporter protein crystal structure and generation of quality data to refine and develop viable predictive models.

Structure based classification for bile salt export pump (BSEP) inhibitors using comparative structural modeling of human BSEP

The bile salt export pump (BSEP) actively transports conjugated monovalent bile acids from the hepatocytes into the bile. This facilitates the formation of micelles and promotes digestion and absorption of dietary fat. Inhibition of BSEP leads to decreased bile flow and accumulation of cytotoxic bile salts in the liver. A number of compounds have been identified to interact with BSEP, which results in drug-induced cholestasis or liver injury. Therefore, in silico approaches for flagging compounds as potential BSEP inhibitors would be of high value in the early stage of the drug discovery pipeline. Up to now, due to the lack of a high-resolution X-ray structure of BSEP, in silico based identification of BSEP inhibitors focused on ligand-based approaches. In this study, we provide a homology model for BSEP, developed using the corrected mouse P-glycoprotein structure (PDB ID: 4M1M). Subsequently, the model was used for docking-based classification of a set of 1212 compounds (405 BSEP inhibitors, 807 non-inhibitors). Using the scoring function ChemScore, a prediction accuracy of 81% on the training set and 73% on two external test sets could be obtained. In addition, the applicability domain of the models was assessed based on Euclidean distance. Further, analysis of the protein–ligand interaction fingerprints revealed certain functional group-amino acid residue interactions that could play a key role for ligand binding. Though ligand-based models, due to their high speed and accuracy, remain the method of choice for classification of BSEP inhibitors, structure-assisted docking models demonstrate reasonably good prediction accuracies while additionally providing information about putative protein–ligand interactions.