Predictive Quantitative Structure-activity Relationship Models Research Articles

QSAR Analysis of Saquinavir Analogues (Peptidomimetics) as Potent HIV‐Protease Inhibitors

The correlation of structural features with the biological activity has always played an important role in drug designing process. The present paper discussesthe 2D‐ and 3D‐ Quantitative structure activity relationship (QSAR) studies, performed on a series of compounds related to saquinavir, an established HIV‐protease inhibitor (PI). The analysis was done on structure based calculations using various methods of QSAR like multiple linear regression (MLR), k‐nearest neighbour (k‐NN) and partial least square (PLS), to establish QSAR models for biological activity prediction of unknown compounds. A total of 27 peptidomimetics (Saquinavir analogues) were used for the study and models were developed using a training set of 22 compounds and test set of 5 compounds. The r2 value of 0.959 and crossvalidated r2 (q2) of 0.926 was obtained when models were generated using physicochemical descriptors during 2D‐QSAR analysis. In case of 3D‐QSAR analysis, database alignment of all compounds was done by field fit of steric and electrostatic molecular fields. 3D‐QSAR models generated showed r2 of 0.81 when steric and electrostatic fields were considered as basis of model generation. The meaningful information obtained from the study can be used for the design of saquinavir analogues having better inhibitory activity for HIV‐protease. Also, the QSAR models generated can be very useful to predict the HIV‐PIs and also for virtual screening for identification of new lead molecules.

In silico development, validation and comparison of predictive QSAR models for lipid peroxidation inhibitory activity of cinnamic acid and caffeic acid derivatives using multiple chemometric and cheminformatics tools

The design and development of antioxidant molecules have lately gained a great deal of focus which is attributed to their immense biomedicinal importance in combating the free radical associated health hazards. In a situation to replenish the endogenous antioxidant loss, synthetic molecules with potent antioxidant activity is demanded. The present work thus aims at in silico modeling of antioxidant molecules that may facilitate in searching and designing of new chemical entities with enhanced activity profile. A series of cinnamic acid and caffeic acid derivatives having the ability to inhibit lipid peroxidation have been modeled in the present work. Three different types of models were developed using different chemometric and cheminformatics tools to identify the essential structural attributes: (a) descriptor based QSAR models, (b) 3D pharmacophore models and (c) HQSAR (hologram QSAR) models. For the conventional QSAR modeling, descriptors belonging to different categories [quantum chemical descriptors (Mulliken charges of the common atoms of the molecules), thermodynamic descriptors, electronic descriptors, structural descriptors and spatial descriptors] were calculated for the development of statistically significant as well as well interpretable quantitative structure-activity relationship (QSAR) models. Two different chemometric tools [genetic function approximation (GFA) and genetic partial least squares (G/PLS)] were employed for the development of the QSAR models. The 3D pharmacophore model focused on the essential pharmacophoric features while the HQSAR model implicated the prime structural fragments that were necessitated for the optimal anti-lipid peroxidative activity of the molecules. All the models were validated based on internal, external and overall validation statistics. Randomization was performed in order to ensure the absence of chance correlation in the developed models. Among all models, the descriptor-based model developed using the GFA-spline technique yielded the most satisfactory results. The results obtained from all the models corroborate well with each other and chiefly signify the importance of the ketonic oxygen of the amide/ acid fragment and the ethereal oxygen substituted on the parent phenyl ring of the molecules under study. Thus the models can efficiently be utilized for extensive screening of large datasets and their subsequent activity prediction.

Development predictive QSAR models for artemisinin analogues by various feature selection methods: A comparative study

Quantitative structure–activity relationship (QSAR) models were derived for 179 analogues of artemisinin, a potent antimalarial agent. The activities of these compounds were investigated by means of multiple linear regression (MLR). To select relevant descriptors, several methods including stepwise selection, successive projection algorithm and an ant colony optimization algorithm (called memorized_ACS) were employed. A wide variety of molecular descriptors belonging to various structural properties were calculated for each molecule. Two matrixes (D1 and D2) of molecular properties were built. The D1 matrix included the calculated descriptors and the D2 matrix contained the first to third orders of the calculated descriptors and the logarithm of absolute values of the calculated descriptors. For both data matrixes, significant QSAR models were obtained by the memorized_ACS algorithm. The reactive and PEOE (partial equalization of orbital electronegativity) descriptors represented the highest impact on the antimalarial activity. The PEOE descriptors belong to partial charge descriptors and the reactive descriptor is an indicator of the presence of the reactive groups in the molecule. The best MLR model has a training error of 0.71 log RA units (r 2 = 0.81) and a prediction error of 0.48 log RA units (r 2 = 0.88).

On Various Metrics Used for Validation of Predictive QSAR Models with Applications in Virtual Screening and Focused Library Design

Quantitative structure-activity relationships (QSARs) have important applications in drug discovery research, environmental fate modeling, property prediction, etc. Validation has been recognized as a very important step for QSAR model development. As one of the important objectives of QSAR modeling is to predict activity/property/toxicity of new chemicals falling within the domain of applicability of the developed models and QSARs are being used for regulatory decisions, checking reliability of the models and confidence of their predictions is a very important aspect, which can be judged during the validation process. One prime application of a statistically significant QSAR model is virtual screening for molecules with improved potency based on the pharmacophoric features and the descriptors appearing in the QSAR model. Validated QSAR models may also be utilized for design of focused libraries which may be subsequently screened for the selection of hits. The present review focuses on various metrics used for validation of predictive QSAR models together with an overview of the application of QSAR models in the fields of virtual screening and focused library design for diverse series of compounds with citation of some recent examples.

Linear and nonlinear quantitative structure–activity relationship modeling of the HIV-1 reverse transcriptase inhibiting activities of thiocarbamates

For a series of thiocarbamates, non-nucleoside HIV-1 reverse transcriptase inhibitors, few descriptors have been selected from a large pool of theoretical molecular descriptors by means of the ant colony optimization (ACO) feature selection method. The selected descriptors were correlated with the bioactivities of the molecules using the well known multiple linear regression (MLR) and partial least squares (PLS) regression techniques, and, to account for nonlinearity, also PLS coupled to radial basis function (RBF) on the one hand and radial basis function neural network (RBFNN) on the other. In this case study, the RBF/PLS results were better than those from the other modeling techniques applied. The prediction ability of the ACO/RBF/PLS-based quantitative structure–activity relationship (QSAR) model was found to be significantly superior to comparative molecular field analysis (CoMFA) and comparative molecular similarity index analysis (CoMSIA) models previously established for this series of compounds. It was also demonstrated that RBF as a nonlinear approach is useful in deriving simple and predictive QSAR models, without the need to recourse to expeditious 3D methodologies.

Quantitative structure–activity relationship analysis of a novel series of chemicals antagonizing WT and MT AR

In the clinical treatment of prostate cancer, mutation in the androgen receptor (AR) that occurred in patients, often results in drug resistance because the agonist activity of the drug increases while the antagonist activity decreases. Recently a novel series of diazacycloundecane AR antagonists were reported, and their abilities to antagonize the wild type (WT) AR as well as antagonize the mutated type (MT) AR (T877A) were determined experimentally. In this paper, to understand more about the drug resistance and to help to design more active antagonists, the quantitative structure–activity relationship (QSAR) of these compounds were analyzed, and the different interactions of these antagonists with two types of AR were compared. As a result two robust and predictive QSAR models were proposed, and the different structure features related with the antagonist activities on WT and MT AR were analyzed separately.

QSAR studies of some substituted imidazolinones angiotensin II receptor antagonists using Partial Least Squares Regression (PLSR) method based feature selection

The Quantitative structure–activity relationship (QSAR) analyses were carried out for a series of imidazolinones as nonpeptide angiotensin II receptor antagonists to find out the structural requirements of their antihypertensive activities. Partial Least Squares Regression (PLSR) method coupled was applied to derive QSAR models which were further validated for statistical significance and predictive ability by internal and external validation. The statistically significant best model with high correlation coefficient (r2 = 0.9031) was selected for further study. The model was further validated by means of crossed squared correlation coefficient (q2 = 0.8652 and pred_r2 = 0.8261) which shows the model has good predictive ability, higher value of F statics 66.831 also validates significance of model, Degree of freedom20. The physicochemical descriptor SssCH2count and alignment-independent descriptors T_N_O_6, T_2_Cl_6 and Chi5 were found to show significant correlation with biologic activity in imidazolinones. Successful implementation of a predictive QSAR model largely depends on the selection of a preferred set of molecular descriptors that can signify the antihypertensive new design molecules.

Predictive QSAR models of 3-acylamino-2-aminopropionic acid derivatives as partial agonists of the glycine site on the NMDA receptor

A quantitative structure–activity relationship (QSAR) model has been carried out for a series of 3-acylamino-2-aminopropionic acid derivatives with high affinities in a binding assay for the glycine site. The replacement method (RM) and stepwise-multiple linear regression (Stepwise-MLR) strategies are used as feature selection (descriptor selection). The model of the relationship between selected molecular descriptors and pK i data has been achieved by linear (multiple linear regression, MLR) and nonlinear (support vector machine, SVM) methods. Leave-one-out cross-validation and external-validation were carried out with the aim for evaluating the predictive capability of the models. The squared correlation coefficients of experimental versus predicted activities for the test set obtained by MLR and SVM models using RM feature selection are 0.463 and 0.655, respectively. Our best QSAR model illustrates the importance of an adequate distribution of atomic properties represented in 3D frames and reveals the mass and van der Waals volumes as the most influencing atomic properties in the structures of the 3-acylamino-2-aminopropionic acid derivatives.

Chembench: a cheminformatics workbench

Advances in the field of cheminformatics have been hindered by a lack of freely available tools. We have created Chembench, a publicly available cheminformatics portal for analyzing experimental chemical structure-activity data. Chembench provides a broad range of tools for data visualization and embeds a rigorous workflow for creating and validating predictive Quantitative Structure-Activity Relationship models and using them for virtual screening of chemical libraries to prioritize the compound selection for drug discovery and/or chemical safety assessment. Freely accessible at: http://chembench.mml.unc.edu alex_tropsha@unc.edu

MIA-QSAR, PCA-ranking and least-squares support-vector machines in the accurate prediction of the activities of phosphodiesterase type 5 (PDE-5) inhibitors

Phosphodiesterase type-5 (PDE-5) is a key enzyme involved in the erection process. PDE-5 inhibitors, such as Sildenafil (ViagraTM), Vardenafil (LevitraTM) and Tadalafil (CialisTM), are used for the treatment of erectile dysfunction. Computer-assisted modelling of biological activities of PDE-5 inhibitors may make quantitative structure–activity relationship (QSAR) models useful for the development of safer (low side effects) and more potent drugs. The multivariate image analysis applied to QSAR (MIA-QSAR) method, coupled to partial least-squares (PLS) regression, has provided highly predictive QSAR models. Nevertheless, regression methods which take into account nonlinearity, such as least-squares support-vector machines (LS-SVMs), are supposed to predict biological activities more accurately than the usual linear methods. Thus, together with prior variable selection using principal component analysis ranking, MIA-QSAR and LS-SVM regression were applied to model the bioactivities of a series of cyclic guanine derivatives (PDE-5 inhibitors), and the results were compared with those based on linear methodologies. MIA-QSAR/LS-SVM was found to improve greatly the prediction performance when compared with MIA-QSAR/PLS, MIA-QSAR/N-PLS, CoMFA/PLS and CoMSIA/PLS models.

Computational quantum chemistry and adaptive ligand modeling in mechanistic QSAR

Drugs are adaptive molecules. They realize this peculiarity by generating different ensembles of prototropic forms and conformers that depend on the environment. Among the impressive amount of available computational drug discovery technologies, quantitative structure-activity relationship approaches that rely on computational quantum chemistry descriptors are the most appropriate to model adaptive drugs. Indeed, computational quantum chemistry descriptors are able to account for the variation of the intramolecular interactions of the training compounds, which reflect their adaptive intermolecular interaction propensities. This enables the development of causative, interpretive and reasonably predictive quantitative structure-activity relationship models, and, hence, sound chemical information finalized to drug design and discovery.

Exploring QSARs with Extended Topochemical Atom (ETA) Indices for Modeling Chemical and Drug Toxicity

Development of quantitative structure-activity relationships (QSARs) and quantitative structure-property relationships (QSPRs) has been practiced for prediction of various toxicities and other relevant properties of chemicals including drug candidates to minimize animal testing, cost and time associated with risk assessment and management processes. This communication reviews published reports of QSARs/QSPRs with Extended Topochemical Atom (ETA) indices for modeling chemical and drug induced toxicities and some physicochemical properties relevant to such toxicities. In each study, ETA models have been compared to those developed using various non-ETA models and it was found that the quality of the QSARs involving ETA parameters were comparable to those involving non-ETA parameters. ETA descriptors were also found to increase statistical quality of the models involving non-ETA parameters when used in combination. On the basis of the reported studies, it can be concluded that the ETA descriptors are sufficiently rich in chemical information to encode the structural features contributing to the toxicities and these indices may be used in combination with other topological and physicochemical descriptors for development of predictive QSAR models. Such models may be used for virtual screening and in silico prediction of toxicities, and if appropriately used, these may be proved helpful for regulatory decision support and decision making processes.

Design, Synthesis, and Evaluation of Antineoplastic Activity of Novel Carbocyclic Nucleosides.

Cancer is the leading cause of death among men and women under age 85. Every year, millions of individuals are diagnosed with cancer. But finding new drugs is a complex, expensive, and very time-consuming task. Over the past decade, the cancer research community has begun to address the in silico modeling approaches, such as Quantitative Structure-Activity Relationships (QSAR), as an important alternative tool for targeting potential anticancer drugs. With the compilation of a large dataset of nucleosides synthesized in our laboratories, or elsewhere, and tested in a single cytotoxic assay under the same experimental conditions, we recognized a unique opportunity to attempt to build predictive QSAR models. Early efforts with 2D classification models built from part of this dataset were very encouraging. Here we report a further detailed evaluation of classification models to flag potential anticancer activities derived from a variety of 3D molecular representations. A quantitative 3D-model model that discriminates anticancer compounds from the inactive ones was attained, which allowed the correct classification of 82 % of compounds in such a large and diverse dataset, with only 5 % of false inactives and 11 % of false actives. The model developed here was then used to select and design a new series of nucleosides, by classifying beforehand them as active/inactive anticancer compounds. From the compounds so designed, 22 were synthesized and evaluated for their inhibitory effects on the proliferation of murine leukemia cells (L1210/0), of which 86 % were well-classified as active or inactive, and only two were false actives, corroborating the good predictive ability of the present discriminant model. The results of this study thus provide a valuable tool for the design of novel potent anticancer nucleoside analogues.

QSAR with quantum topological molecular similarity indices: toxicity of aromatic aldehydes to Tetrahymena pyriformis

Extensive production and utilization of aromatic aldehydes and their derivatives without proper certification is alarming with regard to environmental safety. This concern motivated our construction of predictive quantitative structure–activity relationship (QSAR) models for the toxicity of aldehydes to the ecologically important species Tetrahymena pyriformis. Quantum topological molecular similarity (QTMS) descriptors, along with the lipid-water partition coefficient (log K o/w), were used as predictor variables. The QTMS descriptors were calculated at different levels of theory including AM1, HF/3-21G(d), HF/6-31G(d), B3LYP/6-31 + G(d,p), B3LYP/6-311 + G(2d,p) and MP2/6-311+G(2d,p). The data set of 77 aromatic aldehydes was divided into a training set (n = 58) and a test (n = 19) set, and 58 models were developed using partial least squares (PLS) and genetic partial least squares (G/PLS). We evaluated the overall predictive capacity of the models based on leave-one-out predictions for the training set compounds and model derived predictions for the test set compounds. For both PLS and G/PLS, the models built at the HF/6-31G(d) level show better predictivity (based on overall prediction) than the models developed at any of the other five levels. Further validation was also performed utilizing (process and model) randomization tests. We show that improved predictive QSAR models for aldehydic toxicity to Tetrahymena pyriformis can be generated using QTMS descriptors along with log K o/w.

Comparative QSAR studies of nitrofuranyl amide derivatives using theoretical structural properties

The global increase in multidrug-resistant Mycobacterium tuberculosis strains and intolerance of first-line anti-tuberculosis drugs may cause major health problems and necessitate modification of the structural therapy regimen. In an ongoing effort to develop new and potent anti-tuberculosis agents, a series of nitrofuranyl amides were subjected to quantitative structure–activity relationship (QSAR) analysis using various feature selection methods. Nitrofuranyl amide derivatives with good therapeutic indices are known to inhibit an enzyme responsible for bacterial cell-wall synthesis and act as novel mycobacterial inhibitors. Successful implementation of a predictive QSAR model largely depends on the selection of a preferred set of molecular descriptors that can signify the chemical–biological interaction. Genetic algorithm (GA), simulated annealing and stepwise regression are applied as variable selection methods for an effective comparison and model development. The results of two-dimensional QSAR showed that a combination of topological indices, hydrophobic properties and autocorrelation descriptors of different atomic properties could be explored to design potent anti-tubercular inhibitors. Further analysis using three-dimensional QSAR technique identifies a suitable model obtained by GA-partial least square method leading to anti-tubercular activity prediction. The influences of steric and electrostatic field effects generated by the contribution plot are analysed and discussed. Both two- and three-dimensional QSAR analyses of such derivatives provide important structural insights for designing potent anti-tuberculosis drugs.

On Two Novel Parameters for Validation of Predictive QSAR Models

Validation is a crucial aspect of quantitative structure–activity relationship (QSAR) modeling. The present paper shows that traditionally used validation parameters (leave-one-out Q2 for internal validation and predictive R2 for external validation) may be supplemented with two novel parameters rm2 and Rp2 for a stricter test of validation. The parameter rm2(overall) penalizes a model for large differences between observed and predicted values of the compounds of the whole set (considering both training and test sets) while the parameter Rp2 penalizes model R2 for large differences between determination coefficient of nonrandom model and square of mean correlation coefficient of random models in case of a randomization test. Two other variants of rm2 parameter, rm2(LOO) and rm2(test), penalize a model more strictly than Q2 and R2pred respectively. Three different data sets of moderate to large size have been used to develop multiple models in order to indicate the suitability of the novel parameters in QSAR studies. The results show that in many cases the developed models could satisfy the requirements of conventional parameters (Q2 and R2pred) but fail to achieve the required values for the novel parameters rm2 and Rp2. Moreover, these parameters also help in identifying the best models from among a set of comparable models. Thus, a test for these two parameters is suggested to be a more stringent requirement than the traditional validation parameters to decide acceptability of a predictive QSAR model, especially when a regulatory decision is involved.

Quantitative structure activity relationship and risk analysis of some heavy metal residues in the milk of cattle and goat

Milk of cattle and goat was collected from various localities of Faisalabad, Pakistan. Heavy metal concentration in milk was determined by atomic absorption spectrophotometer. Milk from both of the species was found richly contaminated with cadmium (Cd), chromium (Cr), nickel (Ni), and lead (Pb). Lead concentration in the milk of goat was significantly higher as compared to cattle milk. Quantitative structure activity relationship (QSAR) models were suggested to predict the residues of unknown heavy metals in the milk of cattle and goat using their known physicochemical properties such as molecular weight (MW), melting point (MP), and boiling point (BP) as well as the milk characteristics such as pH, % fat, and specific gravity (SG) of both of these species. The analysis revealed good correlation coefficients (R2 = 0.759) and (R2 = 0.876) for cattle and goat QSAR models, respectively. This analysis indicates the need to construct species specific QSAR models for residue prediction. The coefficients for SG for the studied metals were higher in both cattle and goat milk. This suggests that SG is a better determinant for heavy metal residue prediction in the milk of these animals. Risk analysis was conducted based upon the determined heavy metal residues and their provisional tolerable daily intakes. The daily intake of Cd, Cr, and Pb was found to be 1.6, 27, and 116 times higher, respectively, in cattle milk and 1.8, 29, 262 times higher respectively, in goat milk. This intake of heavy metal contaminated milk might pose health hazards to humans in this locality.

QSAR Models for Prediction of PPARδ Agonistic Activity of Indanylacetic Acid Derivatives

AbstractPeroxisome Proliferator Activated Receptor β/δ (PPAR β/δ), one of three PPAR isoforms is a member of nuclear receptor superfamily and ubiquitously expressed in several metabolically active tissues such as liver, muscle, and fat. Tissue specific expression and knock‐out studies suggest a role of PPARδ in obesity and metabolic syndrome. Specific and selective PPARδ ligands may play an important role in the treatment of metabolic disorders. Indanylacetic acid derivatives reported as potent and specific ligands against PPARδ have been studied for the Quantitative Structure–Activity Relationships (QSAR). Molecules were represented by chemical descriptors that encode constitutional, topological, geometrical, and electronic structure features. Four different approaches, i.e., random selection, hierarchical clustering, k‐means clustering, and sphere exclusion method were used to classify the dataset into training and test subsets. Forward stepwise Multiple Linear Regression (MLR) approach was used to linearly select the subset of descriptors and establish the linear relationship with PPARδ agonistic activity of the molecules. The models were validated internally by Leave One Out (LOO) and externally for the prediction of test sets. The best subset of descriptors was then fed to the Artificial Neural Networks (ANN) to develop non‐linear models. Statistically significant MLR models; with R2 varying from 0.80 to 0.87 were generated based on the different training and test set selection methods. Training of ANNs with different architectures for the different training and test selection methods resulted in models with R2 values varying from 0.83 to 0.94, which indicates the high predictive ability of the models.

Fragment-based QSAR: perspectives in drug design

Drug design is a process driven by innovation and technological breakthroughs involving a combination of advanced experimental and computational methods. A broad variety of medicinal chemistry approaches can be used for the identification of hits, generation of leads, as well as to accelerate the optimization of leads into drug candidates. Quantitative structure-activity relationship (QSAR) methods are among the most important strategies that can be applied for the successful design of small molecule modulators having clinical utility. Hologram QSAR (HQSAR) is a modern 2D fragment-based QSAR method that employs specialized molecular fingerprints. HQSAR can be applied to large data sets of compounds, as well as traditional-size sets, being a versatile tool in drug design. The HQSAR approach has evolved from a classical use in the generation of standard QSAR models for data correlation and prediction into advanced drug design tools for virtual screening and pharmacokinetic property prediction. This paper provides a brief perspective on the evolution and current status of HQSAR, highlighting present challenges and new opportunities in drug design.

Novel Inhibitors of Human Histone Deacetylase (HDAC) Identified by QSAR Modeling of Known Inhibitors, Virtual Screening, and Experimental Validation

Inhibitors of histone deacetylases (HDACIs) have emerged as a new class of drugs for the treatment of human cancers and other diseases because of their effects on cell growth, differentiation, and apoptosis. In this study we have developed several quantitative structure-activity relationship (QSAR) models for 59 chemically diverse histone deacetylase class 1 (HDAC1) inhibitors. The variable selection k nearest neighbor (kNN) and support vector machines (SVM) QSAR modeling approaches using both MolconnZ and MOE chemical descriptors generated from two-dimensional rendering of compounds as chemical graphs have been employed. We have relied on a rigorous model development workflow including the division of the data set into training, test, and external sets and extensive internal and external validation. Highly predictive QSAR models were generated with leave-one-out cross-validated (LOO-CV) q2 and external R2 values as high as 0.80 and 0.87, respectively, using the kNN/MolconnZ approach and 0.93 and 0.87, respectively, using the SVM/MolconnZ approach. All validated QSAR models were employed concurrently for virtual screening (VS) of an in-house compound collection including 9.5 million molecules compiled from the ZINC7.0 database, the World Drug Index (WDI) database, the ASINEX Synergy libraries, and other commercial databases. VS resulted in 45 structurally unique consensus hits that were considered novel putative HDAC1 inhibitors. These computational hits had several novel structural features that were not present in the original data set. Four computational hits with novel scaffolds were tested experimentally, and three of them were confirmed active against HDAC1, with IC50 values for the most active compound of 1.00 microM. The fourth compound was later identified to be a selective inhibitor of HDAC6, a Class II HDAC. Moreover, two of the confirmed hits are marketed drugs, which could potentially facilitate their further development as anticancer agents. This study illustrates the power of the combined QSAR-VS method as a general approach for the effective identification of structurally novel bioactive compounds.