QSAR models in receptor-mediated effects: the nuclear receptor superfamily

Optimization of a Deep-Learning Method Based on the Classification of Images Generated by Parameterized Deep Snap a Novel Molecular-Image-Input Technique for Quantitative Structure-Activity Relationship (QSAR) Analysis

Numerous chemical compounds are distributed around the world and may affect the homeostasis of the endocrine system by disrupting the normal functions of hormone receptors. Although the risks associated with these compounds have been evaluated by acute toxicity testing in mammalian models, the chronic toxicity of many chemicals remains due to high cost of the compounds and the testing, etc. However, computational approaches may be promising alternatives and reduce these evaluations. Recently, deep learning (DL) has been shown to be promising prediction models with high accuracy for recognition of images, speech, signals, and videos since it greatly benefits from large datasets. Recently, a novel DL-based technique called DeepSnap was developed to conduct QSAR analysis using three-dimensional images of chemical structures. It can be used to predict the potential toxicity of many different chemicals to various receptors without extraction of descriptors. DeepSnap has been shown to have a very high capacity in tests using Tox21 quantitative qHTP datasets. Numerous parameters must be adjusted to use the DeepSnap method but they have not been optimized. In this study, the effects of these parameters on the performance of the DL prediction model were evaluated in terms of the loss in validation as an indicator for evaluating the performance of the DL using the toxicity information in the Tox21 qHTP database. The relations of the parameters of DeepSnap such as (1) number of molecules per SDF split into (2) zoom factor percentage, (3) atom size for van der waals percentage, (4) bond radius, (5) minimum bond distance, and (6) bond tolerance, with the validation loss following quadratic function curves, which suggests that optimal thresholds exist to attain the best performance with these prediction models. Using the parameter values set with the best performance, the prediction model of chemical compounds for CAR agonist was built using 64 images, at 105° angle, with AUC of 0.791. Thus, based on these parameters, the proposed DeepSnap-DL approach will be highly reliable and beneficial to establish models to assess the risk associated with various chemicals.

Quantitative structure-activity relationship analysis using deep learning based on a novel molecular image input technique

Quantitative structure-activity relationship (QSAR) analysis uses structural, quantum chemical, and physicochemical features calculated from molecular geometry as explanatory variables predicting physiological activity. Recently, deep learning based on advanced artificial neural networks has demonstrated excellent performance in the discipline of QSAR research. While it has properties of feature representation learning that directly calculate feature values from molecular structure, the use of this potential function is limited in QSAR modeling. The present study applied this function of feature representation learning to QSAR analysis by incorporating 360° images of molecular conformations into deep learning. Accordingly, I successfully constructed a highly versatile identification model for chemical compounds that induce mitochondrial membrane potential disruption with the external validation area under the receiver operating characteristic curve of ≥0.9.

Recent advances in computational chemistry for identification of ligands for biological receptors: interdisciplinary aspects

Background: Computational (in silico) methods, such as quantitative structure-activity relationships (QSARs) are already well recognized and used in many screening programs related to environmental, industrial and medical chemistry. The main idea of the QSAR is that there is a relationship between molecular structure and ultimate biological effect caused by a chemical compound. In this respect the approach could be used successfully for prediction of various biological endpoints caused by chemical compounds including receptor binding affinity. Aim of the study: In the current study the capabilities for structure-activity modelling incorporated in noncommercial software tool have been employed for investigating the binding effect of xenobiotics toward estrogen and human pregnane X receptor. Material and methods: The analysis was performed by making use of the non-commercial software platform QSAR Toolbox. This system allows application of a set of built-in models for different biological effects, and also allows incorporation of new models for other endpoints. Results: Two models have been applied for predicting the binding effect toward estrogen and human pregnane X receptors of a large number of chemicals collected in a single database of high practical concern. The results show that there are many chemicals which are able to bind the investigated receptors. Since those chemicals are encountered in the environment, they could be considered as potential threat for society. Conclusions: The obtained results could be used as initial step for further experimental testing of those chemicals in order to confirm their potential to harm biological systems in the body.

A benzoic acid derivative and flavokawains from Piper species as schistosomiasis vector controls

The search of alternative compounds to control tropical diseases such as schistosomiasis has pointed to secondary metabolites derived from natural sources. Piper species are candidates in strategies to control the transmission of schistosomiasis due to their production of molluscicidal compounds. A new benzoic acid derivative and three flavokawains from Piper diospyrifolium, P. cumanense and P. gaudichaudianum displayed significant activities against Biomphalaria glabrata snails. Additionally, "in silico" studies were performed using docking assays and Molecular Interaction Fields to evaluate the physical-chemical differences among the compounds in order to characterize the observed activities of the test compounds against Biomphalaria glabrata snails.

Structure-activity relationship study on the binding of PBDEs with thyroxine transport proteins

Molecular docking and three-dimensional quantitative structure-activity relationships (3D-QSAR) were used to develop models to predict binding affinity of polybrominated diphenyl ether (PBDE) compounds to the human transthyretin (TTR). Based on the molecular conformations derived from the molecular docking, predictive comparative molecular similarity indices analysis (CoMSIA) models were developed. The results of CoMSIA models were as follows: leave-one-out (LOO) cross-validated squared coefficient q² (LOO) = 0.827 (full model, for all 28 compounds); q² (LOO) = 0.752 (split model, for 22 compounds in the training set); leave-many-out (LMO) cross-validated squared coefficient q² (LMO, two groups) = 0.723 ± 0.100 (full model, for all 28 compounds); q² (LMO, five groups) = 0.795 ± 0.030 (full model, for all 28 compounds); and the predictive squared correlation coefficient r²(pred)  = 0.928 (for six compounds in the test set). The developed CoMSIA models can be used to infer the activities of compounds with similar structural characteristics. In addition, the interaction mechanism between hydroxylated polybrominated diphenyl ethers (HO-PBDEs) and the TTR was explored. Hydrogen bonding with amino acid residues Asp74, Ala29, and Asn27 may be an important determinant for HO-PBDEs binding to TTR. Among them, forming hydrogen bonds with amino acid residues Asp74 might exert a more important function.

Insights into the structural and conformational requirements of polybrominated diphenyl ethers and metabolites as potential estrogens based on molecular docking

PBDEs and their metabolites are of concern due to their increasing concentrations in the environment and their toxic effects. Knowledge about the toxicological mechanisms of PBDEs and metabolites is urgently needed for further screening. The objective of the present study was to explore the structural and conformational requirements of PBDE compounds as human estrogen receptor alpha (hERα) agonists, and further screened out hERα agonists from PBDE compounds. Molecular docking and postdocking analysis were adopted to attain the aim. The obtained results revealed that PBDEs can be primarily screened for their estrogenicity using score values, hydrogen bonds interaction with amino acid residues Glu353 and/or Arg394 might be important for HO-PBDEs' estrogenicity. For most MeO-PBDEs, hydrophobic interaction might be the key factor affecting their estrogenic activity. The current study suggested that molecular docking and postdocking analysis can serve as an efficient pre-screening technique for identifying potential estrogens.

Prediction of toxicity and data exploratory analysis of estrogen-active endocrine disruptors using counter-propagation artificial neural networks

In this work, a novel algorithm for optimization of counter-propagation artificial neural networks has been used for development of quantitative structure-activity relationships model for prediction of the estrogenic activity of endocrine-disrupting chemicals. The search for the best model was performed using genetic algorithms. Genetic algorithms were used not only for selection of the most suitable descriptors for modeling, but also for automatic adjustment of their relative importance. Using our recently developed algorithm for automatic adjustment of the relative importance of the input variables, we have developed simple models with very good generalization performances using only few interpretable descriptors. One of the developed models is in details discussed in this article. The simplicity of the chosen descriptors and their relative importance for this model helped us in performing a detailed data exploratory analysis which gave us an insight in the structural features required for the activity of the estrogenic endocrine-disrupting chemicals.

In silico screening of estrogen-like chemicals based on different nonlinear classification models

Increasing concern is being shown by the scientific community, government regulators, and the public about endocrine-disrupting chemicals that are adversely affecting human and wildlife health through a variety of mechanisms. There is a great need for an effective means of rapidly assessing endocrine-disrupting activity, especially estrogen-simulating activity, because of the large number of such chemicals in the environment. In this study, quantitative structure activity relationship (QSAR) models were developed to quickly and effectively identify possible estrogen-like chemicals based on 232 structurally-diverse chemicals (training set) by using several nonlinear classification methodologies (least-square support vector machine (LS-SVM), counter-propagation artificial neural network (CP-ANN), and k nearest neighbour (kNN)) based on molecular structural descriptors. The models were externally validated by 87 chemicals (prediction set) not included in the training set. All three methods can give satisfactory prediction results both for training and prediction sets, and the most accurate model was obtained by the LS-SVM approach through the comparison of performance. In addition, our model was also applied to about 58,000 discrete organic chemicals; about 76% were predicted not to bind to Estrogen Receptor. The obtained results indicate that the proposed QSAR models are robust, widely applicable and could provide a feasible and practical tool for the rapid screening of potential estrogens.

3D QSAR comparative molecular field analysis on nonsteroidal farnesoid X receptor activators

Three-dimensional quantitative structure-activity relationships (3D QSAR) were performed for a series of farnesoid X receptor activators using comparative molecular field analysis (CoMFA). A training set containing 77 compounds served to establish the models. The best statistical results among all models were obtained with region focusing weighted by a S.D. x coefficient values of 0.8 and a grid spacing of 1.0 (r2=0.963, SEE=0.097; q2=0.742, SEP=0.255). The model was used to predict the potency of 20 test set compounds that were not included in the training set, and the predicted values were in good agreement with the experimental results. The final CoMFA model along with the information obtained from 3D contour maps should be useful for the design of novel FXR ligands having improved potency.

Prediction of estrogenicity: validation of a classification model

(Q)SAR models can be used to reduce animal testing as well as to minimise the testing costs. In particular, classification models have been widely used for estimating endpoints with binary activity. The aim of the present study was to develop and validate a classification-based quantitative structure-activity relationship (QSAR) model for endocrine disruption, based on interpretable mechanistic descriptors related to estrogenic gene activation. The model predicts the presence or absence of estrogenic activity according to a pre-defined cut-off in activity as determined in a recombinant yeast assay. The experimental data was obtained from the literature. A two-descriptor classification model was developed that has the form of a decision tree. The predictivity of the model was evaluated by using an external test set and by taking into account the limitations associated with the applicability domain (AD) of the model. The AD was determined as coverage of the model descriptor space. After removing the compounds present in the training set and the compounds outside of the AD, the overall accuracy of classification of the test chemicals was used to assess the predictivity of the model. In addition, the model was shown to meet the OECD Principles for (Q)SAR Validation, making it potentially useful for regulatory purposes.

Hologram quantitative structure–activity relationships for a series of farnesoid X receptor activators

The farnesoid X receptor (FXR) is an attractive drug target for the development of novel therapeutic agents for the treatment of dyslipidemia and cholestasis. Hologram quantitative structure-activity relationship (HQSAR) studies were conducted on a series of potent FXR activators originated from natural product-like libraries. A training set containing 82 compounds served to establish the models. The best HQSAR model was generated using atoms, bonds, connections, chirality, and donor and acceptor as fragment distinction and fragment size default (4-7) with six components. The model was used to predict the potency of 20 test set compounds that were not included in the training set, and the predicted values were in good agreement with the experimental results. The final HQSAR model and the information obtained from HQSAR 2D contribution maps should be useful for the design of novel FXR ligands having improved potency.

Consensus kNN QSAR: a versatile method for predicting the estrogenic activity of organic compounds in silico. A comparative study with five estrogen receptors and a large, diverse set of ligands

Quantitative structure-activity relationships (QSARs) have proved increasingly useful for predicting the biological activities of molecules (e.g., their binding affinities to different receptors) and can be used in environmental chemistry as a preliminary tool for screening the activities of untested molecules, producing valuable information on which compounds should be tested more thoroughly with experimental affinity assays or in animals. The predictive ability of the consensus kNN QSAR method is corroborated here using a diverse set of 245 compounds, which have been assayed for their relative binding affinities to the estrogen receptor of four species: human (ER alpha and ER beta), calf, mouse, and rat. Leave-one-out cross-validation (LOO-CV) and gamma-randomization tests were applied to the QSAR models for internal validation, and separate training and test sets were used for external validation. The internal predictive abilities of the consensus models for all five data sets were convincing, with cross-validated correlation coefficients (LOO-CV q2 values) varying from 0.69 (human ER beta data) to 0.79 (human ER alpha data). The external predictive abilities were also encouraging, as the predictive r2 scores (pr-r2 values) varied from 0.62 (human ER beta data) to 0.77 (calf and mouse data). The results indicate that consensus kNN QSAR is a feasible method for rapid screening of the estrogenic activity of organic compounds.

Assessment of prediction confidence and domain extrapolation of two structure-activity relationship models for predicting estrogen receptor binding activity

Quantitative structure-activity relationship (QSAR) methods have been widely applied in drug discovery, lead optimization, toxicity prediction, and regulatory decisions. Despite major advances in algorithms and software, QSAR models have inherent limitations associated with a size and chemical-structure diversity of the training set, experimental error, and many characteristics of structure representation and correlation algorithms. Whereas excellent fit to the training data may be readily attainable, often models fail to predict accurately chemicals that are outside their domain of applicability. A QSAR's utility and, in the case of regulatory decisions, justification for usage increasingly depend on the ability to quantify a model's potential for predicting unknown chemicals with some known degree of certainty. It is never possible to predict an unknown chemical with absolute certainty. Here we report on two QSAR models based on different data sets for classification of chemicals according to their ability to bind to the estrogen receptor. The models were developed by using a novel QSAR method, Decision Forest, which combines the results of multiple heterogeneous but comparable Decision Tree models to produce a consensus prediction. We used an extensive cross-validation process to define an applicability domain for model predictions based on two quantitative measures: prediction confidence and domain extrapolation. Together, these measures quantify the accuracy of each prediction within and outside of the training domain. Despite being based on large and diverse training sets, both QSAR models had poor accuracy for chemicals within the domain of low confidence, whereas good accuracy was obtained for those within the domain of high confidence. For prediction in the high confidence domain, accuracy was inversely proportional to the degree of domain extrapolation. The model with a larger training set of 1,092, compared with 232 for the other, was more accurate in predicting chemicals at larger domain extrapolation, and could be particularly useful for rapidly prioritizing potential endocrine disruptors from large chemical universe.

Computational models for predicting the binding affinities of ligands for the wild-type androgen receptor and a mutated variant associated with human prostate cancer

In the present study, values of the binding energy (BE) were calculated for the rat androgen receptor on a data set of 25 steroidal and nonsteroidal compounds for which published values of the observed binding affinity (K(i)) are available. A correlation between BE and pK(i) was evident (r(2) = 0.50) for the entire data set and became more pronounced when the steroids and nonsteroids were plotted separately (r(2) congruent with 0.76). Including BE as an additional descriptor to supplement the default steric-electrostatic descriptors in comparative molecular field analysis dramatically improved the predictive ability of the resulting three-dimensional quantitative structure-activity relationship models. We also demonstrate that the observed loss in ligand specificity between the wild-type (wt) AR and the T877A mutant AR associated with androgen-independent prostate cancer is reflected in decreased BE values (i.e., higher binding affinity) for the antiandrogen pharmaceutical hydroxyflutamide and for several nonandrogenic endogenous steroids, most notably cortisol, corticosterone, 17beta-estradiol, progesterone, and 17alpha-hydroxyprogesterone.

Performance of (consensus) kNN QSAR for predicting estrogenic activity in a large diverse set of organic compounds

A novel method (in the context of quantitative structure-activity relationship (QSAR)) based on the k nearest neighbour (kNN) principle, has recently been introduced for the derivation of predictive structure-activity relationships. Its performance has been tested for estimating the estrogen binding affinity of a diverse set of 142 organic molecules. Highly predictive models have been obtained. Moreover, it has been demonstrated that consensus-type kNN QSAR models, derived from the arithmetic mean of individual QSAR models were statistically robust and provided more accurate predictions than the great majority of the individual QSAR models. Finally, the consensus QSAR method was tested with 3D QSAR and log P data from a widely used steroid benchmark data set.

Spectroscopic QSAR Methods and Self-Organizing Molecular Field Analysis for Relating Molecular Structure and Estrogenic Activity

The performance of three "spectroscopic" quantitative structure-activity relationship (QSAR) methods (eigenvalue (EVA), electronic eigenvalue (EEVA), and comparative spectra analysis (CoSA)) for relating molecular structure and estrogenic activity are critically evaluated. The methods were tested with respect to the relative binding affinities (RBA) of a diverse set of 36 estrogens previously examined in detail by the comparative molecular field analysis method. The CoSA method with (13)C chemical shifts appears to provide a predictive QSAR model for this data set. EEVA (i.e., molecular orbital energy in this context) is a borderline case, whereas the performances of EVA (i.e., vibrational normal mode) and CoSA with (1)H shifts are substandard and only semiquantitative. The CoSA method with (13)C chemical shifts provides an alternative and supplement to conventional 3D QSAR methods for rationalizing and predicting the estrogenic activity of molecules. If CoSA is to be applied to large data sets, however, it is desirable that the chemical shifts are available from common databases or, alternatively, that they can be estimated with sufficient accuracy using fast prediction schemes. Calculations of NMR chemical shifts by quantum mechanical methods, as in this case study, seem to be too time-consuming at this moment, but the situation is changing rapidly. An inherent shortcoming common to all spectroscopic QSAR methods is that they cannot take the chirality of molecules into account, at least as formulated at present. Moreover, the symmetry of molecules may cause additional problems. There are three pairs of enantiomers and nine symmetric (C(2) or C(2)(v)) molecules present in the data set, so that the predictive ability of full 3D QSAR methods is expected to be better than that of spectroscopic methods. This is demonstrated with SOMFA (self-organizing molecular field analysis). In general, the use of external test sets with randomized data is encouraged as a validation tool in QSAR studies.

Comparative molecular field analysis (CoMFA) model using a large diverse set of natural, synthetic and environmental chemicals for binding to the androgen receptor

A large number of natural, synthetic and environmental chemicals are capable of disrupting the endocrine systems of experimental animals, wildlife and humans. These so-called endocrine disrupting chemicals (EDCs), some mimic the functions of the endogenous androgens, have become a concern to the public health. Androgens play an important role in many physiological processes, including the development and maintenance of male sexual characteristics. A common mechanism for androgen to produce both normal and adverse effects is binding to the androgen receptor (AR). In this study, we used Comparative Molecular Field Analysis (CoMFA), a three-dimensional quantitative structure-activity relationship (3D-QSAR) technique, to examine AR-ligand binding affinities. A CoMFA model with r2 = 0.902 and q2 = 0.571 was developed using a large training data set containing 146 structurally diverse natural, synthetic, and environmental chemicals with a 10(6)-fold range of relative binding affinity (RBA). By comparing the binding characteristics derived from the CoMFA contour map with these observed in a human AR crystal structure, we found that the steric and electrostatic properties encoded in this training data set are necessary and sufficient to describe the RBA of AR ligands. Finally, the CoMFA model was challenged with an external test data set; the predicted results were close to the actual values with average difference of 0.637 logRBA. This study demonstrates the utility of this CoMFA model for real-world use in predicting the AR binding affinities of structurally diverse chemicals over a wide RBA range.

Estrogenicity and acute toxicity of selected anilines using a recombinant yeast assay

Suspected estrogen modulators include industrial organic chemicals (i.e., xenoestrogens), and have been shown to consist of alkylphenols, bisphenols, biphenylols, and some hydroxy-substituted polycyclic aromatic hydrocarbons. The most prominent structural feature identified to be important for estrogenic activity is a polar group capable of donating hydrogen bonds (i.e., hydroxyl) on an aromatic system. The present study was undertaken to explore the estrogenic activity and acute toxicity of chemicals containing a weaker hydrogen bond donor group on aromatic systems, i.e., the amino substituent. There is a great deal of chemical similarity between aromatic amines (anilines) and aromatic alcohols (phenols). The chemicals chosen for the current study contained an amino-substituted benzene ring with hydrophobic constituents varying in size and shape. Thus, 37 substituted aromatic amines were assayed for estrogenic activity EC50 and acute toxicity LC50 using the Saccharomyces cerevisiae recombinant yeast assay. While the EC50 of 17-beta-estradiol occurs at the 10(-10) range, the aniline with the greatest activity had an EC50 of 10(-6) M. Thus, anilines, in general, are capable only of very weak estrogenic activity in this assay. A comparison of estrogenic potency between the present group of anilines and a set of previously tested analogous phenols indicated that anilines are consistently less estrogenic than phenols. A comparison of hazard indices (EC50/LC50) of these chemicals revealed that, for the vast majority of anilines, the EC50 and LC50 were in the same order of magnitude. More specifically, estrogenic activity of para-substituted alkylanilines increases with alkyl group size up to 5 carbons in length, after which the acute toxicity of the larger alkyl-substituents precluded the ability of the compound to induce the estrogenic response.