QSAR Methods Research Articles

Computer Aided Drug Design for Multi-Target Drug Design: SAR /QSAR, Molecular Docking and Pharmacophore Methods.

Multi-target drugs against particular multiple targets get better protection, resistance profiles and curative influence by cooperative rules of a key beneficial target with resistance behavior and compensatory elements. Computational techniques can assist us in the efforts to design novel drugs (ligands) with a preferred bioactivity outline and alternative bioactive molecules at an early stage. A number of in silico methods have been explored extensively in order to facilitate the investigation of individual target agents and to propose a selective drug. A different, progressively more significant field which is used to predict the bioactivity of chemical compounds is the data mining method. Some of the previously mentioned methods have been investigated for multi-target drug design (MTDD) to find drug leads interact simultaneously with multiple targets. Several cheminformatics methods and structure-based approaches try to extract information from units working cooperatively in a biomolecular system to fulfill their task. To dominate the difficulties of the experimental specification of ligand-target structures, rational methods, namely molecular docking, SAR and QSAR are vital substitutes to obtain knowledge for each structure in atomic insight. These procedures are logically successful for the prediction of binding affinity and have shown promising potential in facilitating MTDD. Here, we review some of the important features of the multi-target therapeutics discoveries using the computational approach, highlighting the SAR, QSAR, docking and pharmacophore methods to discover interactions between drug-target that could be leveraged for curative benefits. A summary of each, followed by examples of its applications in drug design has been provided. Computational efficiency of each method has been represented according to its main strengths and limitations.

Molecular Modelling Approaches for the Analysis of Histamine Receptors and Their Interaction with Ligands.

Several experimental techniques to analyse histamine receptors are available, e.g. pharmacological characterisation of known or new compounds by different types of assays or mutagenesis studies. To obtain insights into the histamine receptors on a molecular and structural level, crystal structures have to be determined and molecular modelling studies have to be performed. It is widely accepted to generate homology models of the receptor of interest based on an appropriate crystal structure as a template and to refine the resulting models by molecular dynamic simulations. A lot of modelling techniques, e.g. docking, QSAR or interaction fingerprint methods, are used to predict binding modes of ligands and pharmacological data, e.g. affinity or even efficacy. However, within the last years, molecular dynamic simulations got more and more important: First of all, molecular dynamic simulations are very helpful to refine the binding mode of a ligand to a histamine receptor, obtained by docking studies. Furthermore, with increasing computational performance it got possible to simulate complete binding pathways of ions or ligands from the aqueous extracellular phase into the allosteric or orthosteric binding pocket of histamine receptors.

3D-QSAR (CoMFA, CoMSIA) and molecular docking studies of natural flavonoid derivatives to explore structural requirements for MCF-7 cell line inhibition

In the present study, molecular modelling studies have been reported on a series of natural flavonoid derivatives to analyse structure activity relationship studies of MCF-7 inhibitors using comparative molecular field analysis (CoMFA) and comparative molecular similarity indices analysis (CoMSIA)-based QSAR methods. Eighty-four compounds were used as training set to establish the model and 20 compounds were used as external test set to validate these models. The generated models exhibited good statistical results such as q2, r2. CoMFA analysis yielded the q2 of 0.765 and r2 of 0.968 with five components. CoMSIA model generated using steric, electrostatic, hydrophobic, donor and acceptor fields with q2 value of 0.592 and r2 value of 0.932 with six components was found to be the optimal model among the various models generated. According to contour map results, we have recommended the critical sites for modification in structure which will be useful in designing potent compounds with improved activity.

Extreme Gradient Boosting as a Method for Quantitative Structure-Activity Relationships.

In the pharmaceutical industry it is common to generate many QSAR models from training sets containing a large number of molecules and a large number of descriptors. The best QSAR methods are those that can generate the most accurate predictions but that are not overly expensive computationally. In this paper we compare eXtreme Gradient Boosting (XGBoost) to random forest and single-task deep neural nets on 30 in-house data sets. While XGBoost has many adjustable parameters, we can define a set of standard parameters at which XGBoost makes predictions, on the average, better than those of random forest and almost as good as those of deep neural nets. The biggest strength of XGBoost is its speed. Whereas efficient use of random forest requires generating each tree in parallel on a cluster, and deep neural nets are usually run on GPUs, XGBoost can be run on a single CPU in less than a third of the wall-clock time of either of the other methods.

Hologram QSAR study on the electrophoretic mobility of aromatic acids

The quantitative structure property relationship (QSPR) for the electrophoretic mobility of aromatic acids was investigated. Hologram QSAR (HQSAR) method was used to build the QSPR model. The fragment parameters of the developed HQSAR model are fragment size: 2–5; fragment length: 307; fragment distinctions: atoms, bonds and connections. The predictive ability of the developed model was assessed with external test validation and leave-one-out cross validation. The results of the two validations demonstrate that the developed HQSAR model is practicable for describing the quantitative relationship between the hologram information and electrophoretic mobility of the investigated aromatic acids. And the contribution map of the developed HQSAR model can graphically illustrate the influence of molecular structure on the electrophoretic mobility of the investigated aromatic acids. Obviously, HQSAR method is fast, practicable and promising for predicting and studying the electrophoretic mobility of aromatic acids.

Analyzing the substitution effect on the CoMFA results within the framework of density functional theory (DFT)

Though QSAR was originally developed in the context of physical organic chemistry, it has been applied very extensively to chemicals (drugs) which act on biological systems, in this idea one of the most important QSAR methods is the 3D QSAR model. However, due to the complexity of understanding the results it is necessary to postulate new methodologies to highlight their physical-chemical meaning. In this sense, this work postulates new insights to understand the CoMFA results using molecular quantum similarity and chemical reactivity descriptors within the framework of density functional theory. To obtain these insights a simple theoretical scheme involving quantum similarity (overlap, coulomb operators, their euclidean distances) and chemical reactivity descriptors such as chemical potential (μ), hardness (ɳ), softness (S), electrophilicity (ω), and the Fukui functions, was used to understand the substitution effect. In this sense, this methodology can be applied to analyze the biological activity and the stabilization process in the non-covalent interactions on a particular molecular set taking a reference compound.

Molecular Modeling Study on Diazine Indole Acetic Acid Derivatives for CRTH2 Inhibitory Activity.

In the present work, molecular modeling studies have been reported on a series of diazine indole acetic acid derivatives to analyze the structure-activity relationship studies of CRTH2 using fragment (Topomer CoMFA and HQSAR) and field (CoMFA and CoMSIA) based QSAR methods. Twenty-six compounds were used as a training set to establish the model, and six compounds were used as a test set to validate the model. The generated models exhibited good statistical results such as correlation coefficient (r2) and the cross-validated correlation coefficient (q2). Topomer CoMFA analysis yielded the q2 of 0.610 and r2 of 0.981. HQSAR model generated using bond and connectivity as fragment distinction and 3-6 as fragment size has the q2 value of 0.707 and conventional r2 value of 0.892 with five components. CoMFA model was assessed by cross-validated q2 value of 0.543 and r2 value of 0.901 with steric and electrostatic fields. CoMSIA model generated using steric, hydrophobic and donor fields with q2 value of 0.550 and r2 value of 0.888 was found to be the optimal model among the various models generated. The contour maps were generated to analyze the important structural features that regulate their inhibitory potency. From the result of contour maps we have suggested the critical sites for chemical modification which will be useful in designing potent compounds with improved activity.

Application of ligand- and receptor-based approaches for prediction of the HIV-RT inhibitory activity of fullerene derivatives

Fullerene and its derivatives have potential to be utilized in many biomedical applications. In the present study, we investigated the role of fullerene derivatives as inhibitors of HIV-RT by combined protein–ligand docking approach and QSAR methods. The study shows the best predictive QSAR model that represents a two-variable model. It has a good ratio of the number of descriptors and predictive ability. The main contributions to the inhibitory activity are provided by signal JhetZ descriptor and μ (dipole moment, as a measure of the polarity of a compound). The developed GA-MLRA-based model demonstrates a good performance, confirmed by statistics \(\left( {R^{2}_{\text{training}} = 0.867,\;Q^{2} = 0.788,\;R^{2}_{\text{test}} = 0.902} \right).\) The structure–activity analysis of these fullerene analogues allowed us to design and suggest for synthesis a set of new potentially active fullerenes. Finally, the molecular docking analysis was carried out to understand the details of interactions between HIV-RT and fullerene-C60 derivatives.

QSAR Methods to Screen Endocrine Disruptors

The identification of endocrine disrupting chemicals (EDCs) is one of the important goals of environmental chemical hazard screening. We report on in silico methods addressing toxicological studies about EDCs with a special focus on the application of QSAR models for screening purpose. Since Estrogen-like (ER) activity has been extensively studied, the majority of the available models are based on ER-related endpoints. Some of these models are here reviewed and described. As example for their application, we screen an assembled dataset of candidate substitutes for some known EDCs belonging to the chemical classes of phthalates, bisphenols and parabens, selected considering their toxicological relevance and broad application, with the general aim of preliminary assessing their ED potential. The goal of the substitution processes is to advance inherently safer chemicals and products, consistent with the principles of green chemistry. Results suggest that the integration of a family of different models accounting for different endpoints can be a convenient way to describe ED as properly as possible and allow also both to increase the confidence of the predictions and to maximize the probability that most active compounds are correctly found.

3D-QSAR, molecular dynamics simulations and molecular docking studies of benzoxazepine moiety as mTOR inhibitor for the treatment of lung cancer.

According to WHO statistics, lung cancer is one of the leading causes of death among all other types of cancer. Many genes get mutated in lung cancer but involvement of EGFR and KRAS are more common. Unavailability of drugs or resistance to the available drugs is the major problem in the treatment of lung cancer. In the present research, mTOR was selected as an alternative target for the treatment of lung cancer which involves PI3K/AKT/mTOR pathway. 28 synthetic mTOR inhibitors were selected from the literature. Ligand based approach (CoMFA and CoMSIA) and structure based approach (molecular dynamics simulations assisted molecular docking study) were applied for the identification of important features of benzoxazepine moiety, responsible for mTOR inhibition. Three different alignments were tried to obtain best QSAR model, of which, distil was found to be the best method, as it gave good statistical results. In CoMFA, Leave One Out (LOO) cross validated coefficients (q(2)), conventional coefficient (r(2)) and predicted correlation coefficient (r(2)pred) values were found to be 0.615, 0.990 and 0.930, respectively. Similarly in CoMSIA, q(2), r(2)ncv and r(2)pred values were found to be 0.748, 0.986 and 0.933, respectively. Molecular dynamics and simulations study revealed that B-chain of mTOR protein was stable at and above 500 FS with respect to temperature (at and above 298 K), Potential energy (at and above 7669.72 kJ/mol) and kinetic energy (at and above 4009.77 kJ/mol). Molecular docking study was performed on simulated protein of mTOR which helped to correlate interactions of amino acids surrounded to the ligand with contour maps generated by QSAR method. Important features of benzoxazepine were identified by contour maps and molecular docking study which would be useful to design novel molecules as mTOR inhibitors for the treatment of lung cancer.

AMINOMETHANESULPHONIC ACIDS AND ITS N-DERIVATIVES ARE COMPONENTS OF N. GOODS BUFFERS

On the basis of pH measuring the temperature dependence of the second ionization constant (pK2) are set for aminomethansulfonic acid (AMSA) and its N-derivatives. It was found that the area of effective buffering of AMSK and its N-benzyl derivative allows to maintain the pH of acid-compatible with life. The absolute values of the coefficients in the equation pK2 = -Ai+ Bi/T for the aminosulfonic acids correlated with their molar masses. A simultaneous change in values DpH /DT and Bicoefficient for the studied acids QSAR method are stayed. The values of the distribution coefficients in the octanol – water (Pow) and molar refraction are calculated. In the future the data obtained in this study will be used to predict the physical, chemical and toxicological properties of the compounds studied.

Prediction of octanol-air partition coefficients for polychlorinated biphenyls (PCBs) using 3D-QSAR models

Based on the experimental data of octanol-air partition coefficients (KOA) for 19 polychlorinated biphenyl (PCB) congeners, two types of QSAR methods, comparative molecular field analysis (CoMFA) and comparative molecular similarity indices analysis (CoMSIA), are used to establish 3D-QSAR models using the structural parameters as independent variables and using logKOA values as the dependent variable with the Sybyl software to predict the KOA values of the remaining 190 PCB congeners. The whole data set (19 compounds) was divided into a training set (15 compounds) for model generation and a test set (4 compounds) for model validation. As a result, the cross-validation correlation coefficient (q2) obtained by the CoMFA and CoMSIA models (shuffled 12 times) was in the range of 0.825–0.969 (>0.5), the correlation coefficient (r2) obtained was in the range of 0.957–1.000 (>0.9), and the SEP (standard error of prediction) of test set was within the range of 0.070–0.617, indicating that the models were robust and predictive. Randomly selected from a set of models, CoMFA analysis revealed that the corresponding percentages of the variance explained by steric and electrostatic fields were 23.9% and 76.1%, respectively, while CoMSIA analysis by steric, electrostatic and hydrophobic fields were 0.6%, 92.6%, and 6.8%, respectively. The electrostatic field was determined as a primary factor governing the logKOA. The correlation analysis of the relationship between the number of Cl atoms and the average logKOA values of PCBs indicated that logKOA values gradually increased as the number of Cl atoms increased. Simultaneously, related studies on PCB detection in the Arctic and Antarctic areas revealed that higher logKOA values indicate a stronger PCB migration ability. From CoMFA and CoMSIA contour maps, logKOA decreased when substituents possessed electropositive groups at the 2-, 3-, 3′-, 5- and 6- positions, which could reduce the PCB migration ability. These results are expected to be beneficial in predicting logKOA values of PCB homologues and derivatives and in providing a theoretical foundation for further elucidation of the global migration behaviour of PCBs.

Predicting carcinogenicity of organic compounds based on CPDB

Cancer is a major killer of human health and predictions for the carcinogenicity of chemicals are of great importance. In this article, predictive models for the carcinogenicity of organic compounds using QSAR methods for rats and mice were developed based on the data from CPDB. The models was developed based on the data of specific target site liver and classified according to sex of rats and mice. Meanwhile, models were also classified according to whether there is a ring in the molecular structure in order to reduce the diversity of molecular structure. Therefore, eight local models were developed in the final. Taking into account the complexity of carcinogenesis and in order to obtain as much information, DRAGON descriptors were selected as the variables used to develop models. Fitting ability, robustness and predictive power of the models were assessed according to the OECD principles. The external predictive coefficients for validation sets of each model were in the range of 0.711–0.906, and for the whole data in each model were all greater than 0.8, which represents that all models have good predictivity. In order to study the mechanism of carcinogenesis, standardized regression coefficients were calculated for all predictor variables. In addition, the effect of animal sex on carcinogenesis was compared and a trend that female showed stronger tolerance for cancerogen than male in both species was appeared.

The Relative Importance of Domain Applicability Metrics for Estimating Prediction Errors in QSAR Varies with Training Set Diversity.

In QSAR, a statistical model is generated from a training set of molecules (represented by chemical descriptors) and their biological activities (an "activity model"). The aim of the field of domain applicability (DA) is to estimate the uncertainty of prediction of a specific molecule on a specific activity model. A number of DA metrics have been proposed in the literature for this purpose. A quantitative model of the prediction uncertainty (an "error model") can be built using one or more of these metrics. A previous publication from our laboratory ( Sheridan , R. P. J. Chem. Inf. 2013 , 53 , 2837 - 2850 ) suggested that QSAR methods such as random forest could be used to build error models by fitting unsigned prediction errors against DA metrics. The QSAR paradigm contains two useful techniques: descriptor importance can determine which DA metrics are most useful, and cross-validation can be used to tell which subset of DA metrics is sufficient to estimate the unsigned errors. Previously we studied 10 large, diverse data sets and seven DA metrics. For those data sets for which it is possible to build a significant error model from those seven metrics, only two metrics were sufficient to account for almost all of the information in the error model. These were TREE_SD (the variation of prediction among random forest trees) and PREDICTED (the predicted activity itself). In this paper we show that when data sets are less diverse, as for example in QSAR models of molecules in a single chemical series, these two DA metrics become less important in explaining prediction error, and the DA metric SIMILARITYNEAREST1 (the similarity of the molecule being predicted to the closest training set compound) becomes more important. Our recommendation is that when the mean pairwise similarity (measured with the Carhart AP descriptor and the Dice similarity index) within a QSAR training set is less than 0.5, one can use only TREE_SD, PREDICTED to form the error model, but otherwise one should use TREE_SD, PREDICTED, SIMILARITYNEAREST1.

CoRILISA: a local similarity based receptor dependent QSAR method.

Molecular similarity methods have played a crucial role in the success of structure-based and computer-assisted drug design. However, with the exception of CoMSIA, the current approaches for estimating molecular similarity yield a global picture thereby providing limited information about the local spatial molecular features responsible for the variation of activity with the 3D structure. Application of molecular similarity measures, each related to the functional "pieces" of a ligand-receptor complex, is advantageous over a composite molecular similarity alone and will provide more insights to rationally interpret the activity based on the receptor and ligand structural features. Building on the ideas of our previously published methodologies-CoRIA and LISA, we present here a local molecular similarity based receptor dependent QSAR method termed CoRILISA which is a hybrid of the two approaches. The method improves on previous techniques by inclusion of receptor attributes for the calculation and comparison of similarity between molecules. For validation studies, the CoRILISA methodology was applied on three large and diverse data sets-glycogen phosphorylase b (GPb), human immunodeficiency virus-1 protease (HIV PR), and cyclin dependent kinase 2 (CDK2) inhibitors. The statistics of the CoRILISA models were benchmarked against the standard CoRIA approach and with other published approaches. The CoRILISA models were found to be significantly better, especially in terms of the predictivity for the test set. CoRILISA is able to identify the thermodynamic properties associated with residues that define the active site and modulate the variation in the activity of the molecules. It is a useful tool in the fragment-based drug discovery approach for ligand activity prediction.

Combined pharmacophore modeling, 3D-QSAR, homology modeling and docking studies on CYP11B1 inhibitors.

The mitochondrial cytochrome P450 enzymes inhibitor steroid 11β-hydroxylase (CYP11B1) can decrease the production of cortisol. Therefore, these inhibitors have an effect in the treatment of Cushing’s syndrome. A pharmacophore model generated by Genetic Algorithm with Linear Assignment for Hypermolecular Alignment of Datasets (GALAHAD) was used to align the compounds and perform comparative molecular field analysis (CoMFA) with Q2 = 0.658, R2 = 0.959. The pharmacophore model contained six hydrophobic regions and one acceptor atom, and electropositive and bulky substituents would be tolerated at the A and B sites, respectively. A three-dimensional quantitative structure-activity relationship (3D-QSAR) study based on the alignment with the atom root mean square (RMS) was applied using comparative molecular field analysis (CoMFA) with Q2 = 0.666, R2 = 0.978, and comparative molecular similarity indices analysis (CoMSIA) with Q2 = 0.721, R2 = 0.972. These results proved that all the models have good predictability of the bioactivities of inhibitors. Furthermore, the QSAR models indicated that a hydrogen bond acceptor substituent would be disfavored at the A and B groups, while hydrophobic groups would be favored at the B site. The three-dimensional (3D) model of the CYP11B1 was generated based on the crystal structure of the CYP11B2 (PDB code 4DVQ). In order to probe the ligand-binding modes, Surflex-dock was employed to dock CYP11B1 inhibitory compounds into the active site of the receptor. The docking result showed that the imidazolidine ring of CYP11B1 inhibitors form H bonds with the amino group of residue Arg155 and Arg519, which suggested that an electronegative substituent at these positions could enhance the activities of compounds. All the models generated by GALAHAD QSAR and Docking methods provide guidance about how to design novel and potential drugs for Cushing’s syndrome treatment.

Ligand biological activity predictions using fingerprint-based artificial neural networks (FANN-QSAR).

This chapter focuses on the fingerprint-based artificial neural networks QSAR (FANN-QSAR) approach to predict biological activities of structurally diverse compounds. Three types of fingerprints, namely ECFP6, FP2, and MACCS, were used as inputs to train the FANN-QSAR models. The results were benchmarked against known 2D and 3D QSAR methods, and the derived models were used to predict cannabinoid (CB) ligand binding activities as a case study. In addition, the FANN-QSAR model was used as a virtual screening tool to search a large NCI compound database for lead cannabinoid compounds. We discovered several compounds with good CB2 binding affinities ranging from 6.70 nM to 3.75 μM. The studies proved that the FANN-QSAR method is a useful approach to predict bioactivities or properties of ligands and to find novel lead compounds for drug discovery research.

In silico machine learning methods in drug development.

Machine learning (ML) computational methods for predicting compounds with pharmacological activity, specific pharmacodynamic and ADMET (absorption, distribution, metabolism, excretion and toxicity) properties are being increasingly applied in drug discovery and evaluation. Recently, machine learning techniques such as artificial neural networks, support vector machines and genetic programming have been explored for predicting inhibitors, antagonists, blockers, agonists, activators and substrates of proteins related to specific therapeutic targets. These methods are particularly useful for screening compound libraries of diverse chemical structures, "noisy" and high-dimensional data to complement QSAR methods, and in cases of unavailable receptor 3D structure to complement structure-based methods. A variety of studies have demonstrated the potential of machine-learning methods for predicting compounds as potential drug candidates. The present review is intended to give an overview of the strategies and current progress in using machine learning methods for drug design and the potential of the respective model development tools. We also regard a number of applications of the machine learning algorithms based on common classes of diseases.

In silico study combining docking and QSAR methods on a series of matrix metalloproteinase 13 inhibitors.

Matrix metalloproteinase 13 (MMP-13) plays an important role in the degradation of articular cartilage and has been considered as an attractive target for the treatment of osteoarthritis; hence, the development of efficient inhibitors of MMP-13 has become a hot study field. Taking a series of carboxylic acid-based MMP-13 inhibitors as research object, this work utilized an extended QSAR method to analyze the structure-activity relationships. We focused on two important topics in QSAR: bioactive conformation and descriptors. Firstly, molecular docking was carried out to dock all molecules into the MMP-13 active site in order to obtain the bioactive conformation. Secondly, based on the docked complex, descriptors characterizing receptor-ligand interactions and the ligand structure were calculated. Thirdly, a genetic algorithm (GA) and multiple linear regression (MLR) were employed to select important descriptors related to inhibitory activities, simultaneously, to build the predictive model. The built model gave satisfactory results with highly accurate fitting and strong external predictive abilities for chemicals not used in model development. Furthermore, the selected descriptors were explored to elucidate important factors influencing the inhibition activities. This study demonstrates that the selection strategy of the docking-guided bioactive conformation is rational and useful in predicting MMP-13 inhibitor activities, and receptor-ligand complex descriptors have an advantage over directly reflecting receptor-ligand interactions.

Identification of novel peroxisome proliferator-activated receptor-gamma (PPARγ) agonists using molecular modeling method.

Peroxisome proliferator-activated receptor-gamma (PPARγ) plays a critical role in lipid and glucose homeostasis. It is the target of many drug discovery studies, because of its role in various disease states including diabetes and cancer. Thiazolidinediones, a synthetic class of agents that work by activation of PPARγ, have been used extensively as insulin-sensitizers for the management of type 2 diabetes. In this study, a combination of QSAR and docking methods were utilised to perform virtual screening of more than 25 million compounds in the ZINC library. The QSAR model was developed using 1,517 compounds and it identified 42,378 potential PPARγ agonists from the ZINC library, and 10,000 of these were selected for docking with PPARγ based on their diversity. Several steps were used to refine the docking results, and finally 30 potentially highly active ligands were identified. Four compounds were subsequently tested for their in vitro activity, and one compound was found to have a K i values of <5 μM.