QSAR by LFER model of HIV protease inhibitor mannitol derivatives using FA-MLR, PCRA, and PLS techniques

Azole Compounds as Inhibitors of Candida albicans: QSAR Modelling

Candida albicans is a pathogenic opportunistic yeast found in the human gut flora. It may also live outside of the human body, causing diseases ranging from minor to deadly. Candida albicans begins as a budding yeast that can become hyphae in response to a variety of environmental or biological triggers. The hyphae form is responsible for the development of multidrug resistant biofilms, despite the fact that both forms have been associated to virulence Here, we have proposed a linear and SPA-linear quantitative structure activity relationship (QSAR) modeling and prediction of Candida albicans inhibitors. A data set that consisted of 60 derivatives of benzoxazoles, benzimidazoles, oxazolo (4, 5-b) pyridines have been used. In this study, that after applying the leverage analysis method to detect outliers' molecules, the total number of these compounds reached 55. SPA-MLR model shows superiority over the multiple linear regressions (MLR) by accounting 90% of the Q ² of anti-fungus derivatives 'activity. This paper focuses on investigating the role of SPA-MLR in developing model. The accuracy of SPA-MLR model was illustrated using leave-one-out (LOO). The mean effect of descriptors and sensitivity analysis show that RDF090u is the most important parameter affecting the as behavior of the inhibitors of Candida albicans.

A natural language processing approach based on embedding deep learning from heterogeneous compounds for quantitative structure-activity relationship modeling

Over the past decade, rapid development in biological and chemical technologies such as high-throughput screening, parallel synthesis, has been significantly increased the amount of data, which requires the creation and the integration of new analytical methods, especially deep learning models. Recently, there is an increasing interest in deep learning utilization in computer-aided drug discovery due to its exceptional successful application in many fields. The present work proposed a natural language processing approach, based on embedding deep neural networks. Our method aims to transform the Simplified Molecular Input Line Entry System format into word embedding vectors to represent the semantics of compounds. These vectors are fed into supervised machine learning algorithms such as convolutional long short-term memory neural network, support vector machine, and random forest to build up quantitative structure-activity relationship models on toxicity data sets. The obtained results on toxicity data to the ciliate Tetrahymena pyriformis (IGC₅₀ ), and acute toxicity rat data expressed as median lethal dose of treated rats (LD₅₀ ) show that our approach can eventually be used to predict the activities of chemical compounds efficiently. All material used in this study is available online through the GitHub portal (https://github.com/BoukeliaAbdelbasset/NLPDeepQSAR.git).

Current approaches for choosing feature selection and learning algorithms in quantitative structure-activity relationships (QSAR)

INTRODUCTION

Quantitative structure-activity/property relationships (QSAR/QSPR) are statistical models which quantitatively correlate quantitative chemical structure information (described as molecular descriptors) to the response end points (biological activity, property, toxicity, etc.). Important strategies for QSAR model development and validation include dataset curation, variable selection, and dataset division, selection of modeling algorithms and appropriate measures of model validation. Areas covered: Different feature selection methods and various linear and nonlinear learning algorithms are employed to address the complexity of data sets for selection of appropriate features important for the responses being modeled, to reduce overfitting of the models, and to derive interpretable models. This review provides an overview of various feature selection methods as well as different statistical learning algorithms for QSAR modeling at an elementary level for nonexpert readers. Expert opinion: Novel sets of descriptors are being continuously introduced to this field; therefore, to handle this issue, there is a need to improve new tools for feature selection, which can lead to development of statistically meaningful models, usable by nonexperts in the fields. While handling data sets of limited size, special techniques like double cross-validation and consensus modeling might be more meaningful in order to remove the possibility of bias in descriptor selection.

An Introduction to the Basic Concepts in QSAR-Aided Drug Design

The need for the development of new drugs to combat existing and newly identified conditions is unavoidable. One of the important tools used in the advanced drug development pipeline is computer-aided drug design. Traditionally, to find a drug many ligands were synthesized and evaluated for their effectiveness using suitable bioassays and if all other drug-likeness features were met, the candidate(s) would possibly reach the market. Although this approach is still in use in advanced format, computational methods are an indispensable component of modern drug development projects. One of the methods used from very early days of rationalizing the drug design approaches is Quantitative Structure-Activity Relationship (QSAR). This chapter overviews QSAR modeling steps by introducing molecular descriptors, mathematical model development for relating biological activities to molecular structures, and model validation. At the end, several successful cases where QSAR studies were used extensively are presented.

An Introduction to the Basic Concepts in QSAR-Aided Drug Design

The need for the development of new drugs to combat existing and newly identified conditions is unavoidable. One of the important tools used in the advanced drug development pipeline is computer-aided drug design. Traditionally, to find a drug many ligands were synthesized and evaluated for their effectiveness using suitable bioassays and if all other drug-likeness features were met, the candidate(s) would possibly reach the market. Although this approach is still in use in advanced format, computational methods are an indispensable component of modern drug development projects. One of the methods used from very early days of rationalizing the drug design approaches is Quantitative Structure-Activity Relationship (QSAR). This chapter overviews QSAR modeling steps by introducing molecular descriptors, mathematical model development for relating biological activities to molecular structures, and model validation. At the end, several successful cases where QSAR studies were used extensively are presented.

Revealing interaction mode between HIV-1 protease and mannitol analog inhibitor

HIV protease is a key enzyme to play a key role in the HIV-1 replication cycle and control the maturation from HIV viruses to an infectious virion. HIV-1 protease has become an important target for anti-HIV-1 drug development. Here, we used molecular dynamics simulation to study the binding mode between mannitol derivatives and HIV-1 protease. The results suggest that the most active compound (M35) has more stable hydrogen bonds and stable native contacts than the less active one (M17). These mannitol derivatives might have similar interaction mode with HIV-1 protease. Then, 3D-QSAR was used to construct quantitative structure-activity models. The cross-validated q(2) values are found as 0.728 and 0.611 for CoMFA and CoMSIA, respectively. And the non-cross-validated r(2) values are 0.973 and 0.950. Nine test set compounds validate the model. The results show that this model possesses better prediction ability than the previous work. This model can be used to design new chemical entities and make quantitative prediction of the bioactivities for HIV-1 protease inhibitors before resorting to in vitro and in vivo experiment.

Designing hypothesis of substituted benzoxazinones as HIV-1 reverse transcriptase inhibitors: QSAR approach

A linear quantitative structure activity relationship (QSAR) model is presented for predicting human immunodeficiency virus-1 (HIV-1) reverse transcriptase enzyme inhibition. The 2D QSAR and 3D-QSAR models were developed by stepwise multiple linear regression, partial least square (PLS) regression and k-nearest neighbor-molecular field analysis, PLS regression, respectively using a database consisting of 33 recently discovered benzoxazinones. The primary findings of this study is that the number of hydrogen atoms, number of (-NH2) group connected with solitary single bond alters the inhibition of HIV-1 reverse transcriptase. Further, presence of electrostatic, hydrophobic and steric field descriptors significantly affects the ability of benzoxazinone derivatives to inhibit HIV-1 reverse transcriptase. The selected descriptors could serve as a primer for the design of novel and potent antagonists of HIV-1 reverse transcriptase.

Modeling the activity of 2-phenylnaphthalene inhibitors using self-training artificial neural networks

The present study investigates the quantitative structure-activity relationship (QSAR) of 2-phenylnaphthalene ligands on an estrogen receptor (ERα). A data set comprising 70 derivatives of 2-phenylnaphthalene is used. The most suitable parameters, classified as topological, geometric and electronic are selected using a combination of genetic algorithm and multiple linear regression (GA-MLR) methods. Then, selected descriptors are used as inputs for a self-training artificial neural network (STANN). Analysis of the results suggests that the STANN model shows superior results compared to the multiple linear regressions (MLR) by accounting for 91.0% of the variances of the antiseptic potency of the 2-phenylnaphthalene derivatives. The accuracy of the 8-4-1 STANN model is illustrated using leave-multiple-out (LMO) cross-validation and Y-randomization techniques.

Exploring structural requirements of 1-N-substituted thiocarbamoyl-3-phenyl-2-pyrazolines as antiamoebic agents using comparative QSAR modelling

Amoebiasis is a potentially lethal disease and causes 70,000 deaths per year. To find structural requirements for more active antiamoebic agents than metronidazole, comparative QSAR modelling was done on thirty 1-N-substituted thiocarbamoyl-3-phenyl-2-pyrazolines. The best model was obtained by using PLS technique with R(A)(2) and R(CV)(2) value of 88.50% and 82.90%, respectively. Amoebicidal activity may increase when Wang-Ford charges at atom numbers 6 and 12 have large positive values. Number of six-membered ring and sum of Kier-Hall electrotopological states may also increase amoebicidal activity when these have large positive values. Increasing value of rotatable bond fraction, approximate surface area and mean atomic polarizability scaled on carbon atom may be detrimental for antiamoebic activity. Decrease in values of electrostatic potential charges at atom numbers 1 and 12 may be conducive for activity. Electrophilic attacks may be favourable at these positions.

Predictive comparative QSAR modelling of (phenylpiperazinyl-alkyl) oxindoles as selective 5-HT1A antagonists by stepwise regression, PCRA, FA-MLR and PLS techniques

5-Hydroxytryptamine, a neurotransmitter released by 5-HT neurons in raphe nuclei and 5-HT(1A) receptors are involved in the pain mechanism of migraine, prevention of postpartum haemorrhage, CNS effects like sleep, anxiety and thermoregulation. Comparative QSAR study was done on thirtytwo (phenylpiperazinyl-alkyl) oxindoles using stepwise regression, PCRA, FA-MLR and PLS techniques to find structurally significant models. ETSA indices at atom numbers 19, 20 and 22, RTSA indices at atom numbers 6, 10 and 20, charge at atom number 19 and presence of chlorine at the atom number 6 may be conducive for the receptor inhibition. Electrophilic attack at atom number 21 may be unfavourable but nucleophilic attack at atom numbers 8 and 14 may be beneficial for % 5-HT(1A) inhibition.

Small-sized human immunodeficiency virus type-1 protease inhibitors containing allophenylnorstatine to explore the S2' pocket

A series of HIV protease inhibitor based on the allophenylnorstatine structure with various P(2)' moieties were synthesized. Among these analogues, we discovered that a small allyl group would maintain potent enzyme inhibitory activity compared to the o-methylbenzyl moiety in clinical candidate 1 (KNI-764, also known as JE-2147, AG-1776, or SM-319777). Introduction of an anilinic amino group to 2 (KNI-727) improved water-solubility and anti-HIV-1 activity. X-ray crystallographic analysis of 13k (KNI-1689) with a beta-methallyl group at P(2)' position revealed hydrophobic interactions with Ala28, Ile84, and Ile50' similar to that of 1. The presence of an additional methyl group on the allyl group in compound 13k significantly increased anti-HIV activity over 1 while providing a rational drug design for structural minimization and improving membrane permeability.

Modeling the antileishmanial activity screening of 5-nitro-2-heterocyclic benzylidene hydrazides using different chemometrics methods

QSAR analysis for modeling the antileishmanial activity screening of a series of 49 nitro derivatives of Hydrazides were carried out using different Chemometrics methods. First, a large number of descriptors were calculated using Hyperchem, Mopac and Dragon softwares. Then, a suitable number of these descriptors were selected using multiple linear regression (MLR) technique. Then selected descriptors were used as inputs for artificial neural networks with three different weight update functions including Levenberg-Marquardt back propagation network (LM-ANN), resilient back propagation network (RP-ANN) and variable learning rate algorithm (GDX-ANN). The best artificial neural network model was an LM-ANN with a 5-5-1 architecture. Comparison of the results indicates that the LM-ANN method has better predictive power than the other methods.

Quantitative structure-activity relationship studies on 1-aryl-tetrahydroisoquinoline analogs as active anti-HIV agents

Predictive quantitative structure-activity relationship analysis was developed for a diverse series of recently synthesized 1-aryl-tetrahydroisoquinoline analogs with anti-HIV activities in this study. The conventional 2D-QSAR models were developed by genetic function approximation (GFA) and stepwise multiple linear regression (MLR) with acceptable explanation of 94.9% and 95.5% and good predicted power of 91.7% and 91.7%, respectively. The results of the 2D-QSAR models were further compared with 3D-QSAR model generated by molecular field analysis (MFA), investigating the substitutional requirements for the favorable receptor-drug interaction and quantitatively indicating the important regions of molecules for their activities. The results obtained by combining these methodologies give insights into the key features for designing more potent analogs against HIV.

Computer-aided prediction for medicinal chemistry via the Internet

Some computational tools for medicinal chemistry freely available on the Internet were compared to examine whether the results of prediction obtained with different methods coincided or not. It was shown that the correlation coefficients varied from 0.65 to 0.90 for log P (seven methods), from 0.01 to 0.73 for aqueous solubility (four methods), and from 0.19 to 0.73 for drug-likeness (three methods). While for log P estimates, reasonable average pairwise correlation was found, for aqueous solubility and drug-likeness it was rather poor. Therefore, using computational tools freely available via the Internet, medicinal chemists should evaluate their accuracy versus experimental data for particular series of compounds. In contrast to prediction of above mentioned properties, which can be done with several Internet tools, wide profiling of biological activity can be obtained only with PASS Inet (http://www.ibmc.msk.ru/PASS). PASS Inet was tested by a dozen medicinal chemists for compounds from different chemical series with various kinds of biological activity, and in the majority of cases the results of prediction coincided with the experiments. New anxiolytics, antiarrhythmics, antileishmanials, and other biologically active agents have been identified on this basis. The advantages and limitations of computer-aided predictions for medicinal chemistry via the Internet are discussed.

Exploring molecular shape analysis of styrylquinoline derivatives as HIV-1 integrase inhibitors

HIV-1 integrase inhibitory activity data of styrylquinoline derivatives have been subjected to 3D-QSAR study by molecular shape analysis (MSA) technique using Cerius(2) version 4.8 software (Accelrys). For the selection of test set compounds, initially a QSAR analysis was done based on topological and structural descriptors and K-means clustering technique was used to classify the entire data set (n=36). Clusters were formed from the factor scores of the whole data set comprising of topological and structural descriptors without the biological activity, and based on the clusters, the data set was divided into training and test sets (n=26 and n=10, respectively) so that all clusters are properly represented in both training and test sets. In the molecular shape analysis, the major steps were (1) generation of conformers and energy minimization; (2) hypothesizing an active conformer (global minimum of the most active compound); (3) selecting a candidate shape reference compound (based on active conformation); (4) performing pair-wise molecular superimposition using maximum common subgroup [MCSG] method; (5) measuring molecular shape commonality using MSA descriptors; (6) determination of other molecular features by calculating spatial and conformational parameters; (7) selection of conformers; (8) generation of QSAR equations by standard statistical techniques. The best model obtained from stepwise regression and GFA techniques shows 51.6% predicted variance (leave-one-out) and 57.3% explained variance. In case of FA-PLS regression, the best relation shows 54.0% predicted variance and 57.9% explained variance. The R(2)(pred) and R(2)(test) values for the GFA derived model are 0.611 and 0.664, respectively, while the best FA-PLS model has R(2)(pred) and R(2)(test) values of 0.602 and 0.656, respectively. These models show the importance of Jurs descriptors (total polar surface area, relative polar surface area, relative hydrophobic surface area, relative positive charge), fraction area of the molecular shadow in the XZ plane (ShadowXZfrac), common overlap steric volume and the ratio of common overlap steric volume to volume of individual molecules. Statistically reliable MSA models obtained from this study suggest that this technique could be useful to design potent HIV-1 integrase inhibitors.

Structural finding of R/S-3,4-dihydro-2,2-dimethyl-6-halo-4-(substituted phenylaminocarbonylamino)-2H-1-benzopyrans as selective pancreatic β-cells KATP-pβ channel openers

R/S-3,4-Dihydro-2,2-dimethyl-6-halo-4-(substituted phenylaminocarbonyl-amino)-2H-1-benzopyrans are pancreatic β-cells potassium (KATP-pβ) channel openers with inhibitory effect on insulin secretion. To find the more active and effective benzopyrans as selective potassium (KATP-pβ) channel openers towards the pancreatic tissues, quantitative structure–activity relationships (QSAR) study was performed using E-state and R-state indices along with Wang–Ford charges, n-octanol/water partition coefficient, molar refractivity, and indicator parameters. QSAR models were developed by statistical techniques, e.g., multiple linear regression (MLR), principle component regression analysis (PCRA), and partial least squares (PLS) analysis. The generated equations were validated by the leave-one-out cross-validation method. The models show the importance of ETSA indices of atom numbers 16, 17, 18, 19, 21 as well as 22. The positive coefficient of S16, S17, S18, S19, S21, and S22 indicate that with the increase of the value of E-state indices, desired activity decreases. RTSA index is also important for the biological activity, and the atom numbers 16, 17, 18, 19, 20 and 22 are involved in van der Waals interactions. RTSA index also possesses negative impact on the inhibition of residual insulin secretion. Wang–Ford charges of some particular atoms are also important for the inhibition. Increase of n-octanol/water partition coefficients of compounds inhibit insulin secretion, and the presence of chlorine atom at m- and p- positions of the phenyl ring B is necessary for the inhibition of residual insulin secretion.Key words: benzopyran derivatives, potassium channel openers, PCRA, PLS, QSAR.

Prediction of cytotoxicity data (CC(50)) of anti-HIV 5-phenyl-1-phenylamino-1H-imidazole derivatives by artificial neural network trained with Levenberg-Marquardt algorithm

A Levenberg-Marquardt algorithm trained feed-forward artificial neural network in quantitative structure-activity relationship (QSAR) was developed for modeling of cytotoxicity data for anti-HIV 5-phenyl-1-phenylamino-1H-imidazole derivatives. A large number of descriptors were calculated with Dragon software and a subset of calculated descriptors was selected with a stepwise regression as a feature selection technique. The 28 molecular descriptors selected by stepwise regression, as the most feasible descriptors, were used as inputs for feed-forward neural network. The neural network architecture and its parameters were optimized. The data were randomly divided into 31 training and 11 validation sets. The prediction ability of the model was evaluated using validation data set and "one-leave-out" cross validation method. The root mean square errors (RMSE) and mean absolute errors for the validation data set were 0.042 and 0.024, respectively. The prediction ability of ANN model was also statistically compared with results of linear free energy related model. The obtained results show the validity of proposed model in the prediction of cytotoxicity data of corresponding anti-HIV drugs.

Quantitative structure-activity relationship study on the biodegradation of acid dyestuffs

Quantitative structure-biodegradability relationships (QSBRs) were established to develop predictive models and mechanistic explanations for acid dyestuffs as well as biological activities. With a total of four descriptors, molecular weight (M(W)), energies of the highest occupied molecular orbital (E(HOMO)), the lowest unoccupied molecular orbital (E(LUMO)), and the excited state (E(ES)), calculated using quantum chemical semi-empirical methodology, a series of models were analyzed between the dye biodegradability and each descriptor. Results showed that E(HOMO) and M(W) were the dominant parameters controlling the biodegradability of acid dyes. A statistically robust QSBR model was developed for all studied dyes, with the combined application of E(HOMO) and M(W). The calculated biodegradations fitted well with the experimental data monitored in a facultative-aerobic process, indicative of the reliable prediction and mechanistic character of the developed model.

Comparative QSAR modeling of CCR5 receptor binding affinity of substituted 1-(3,3-diphenylpropyl)-piperidinyl amides and ureas

The present QSAR study attempts to explore the structural and physicochemical requirements of substituted 1-(3,3-diphenylpropyl)-piperidinyl amides and ureas for CCR5 binding affinity using linear free energy-related (LFER) model of Hansch. QSAR models have been developed using electronic (Hammett sigma), hydrophobicity (pi), and steric (molar refractivity and STERIMOL L, B1, and B5) parameters of phenyl ring substituents of the compounds along with appropriate dummy variables. Whole molecular descriptor like partition coefficient (logP(calcd)) was also tried as an additional descriptor. Statistical techniques like stepwise regression, multiple linear regression with factor analysis as the data preprocessing step (FA-MLR), partial least squares with factor analysis as the preprocessing step (FA-PLS), principal component regression analysis (PCRA), multiple linear regression with genetic function approximation (GFA-MLR), and genetic partial least squares (G/PLS) were applied to identify the structural and physicochemical requirements for the CCR5 binding affinity. The generated equations were statistically validated using leave-one-out technique. The quality of equations obtained from stepwise regression, FA-MLR, FA-PLS, and PCRA is of acceptable statistical range (explained variance ranging from 71.9% to 80.4%, while predicted variance ranging from 67.4% to 77.0%). The GFA-derived models show high intercorrelation among predictor variables used in the equations while the G/PLS model shows lowest statistical quality among all types of models. The best models were also subjected to leave-25%-out crossvalidation.