Toward a general predictive QSAR model for gamma-secretase inhibitors.

Gamma secretase (GS) is an appealing drug target for Alzheimer disease and cancer because of its central role in the processing of amyloid precursor protein and the notch family of proteins. In the absence of three-dimensional structure of GS, there is an urgent need for new methods for the prediction and screening of GS inhibitors, for facilitating discovery of novel GS inhibitors. The present study reports QSAR studies on diverse chemical classes comprising 233 compounds collected from the ChEMBL database. Herein, continuous [PLS regression and neural-network (NN)] and categorical QSAR models (NN and linear discriminant analysis) were developed to obtain pertinent descriptors responsible for variation of GS inhibitor potency. Also, SAR within various chemical classes of compounds is analyzed with respect to important QSAR descriptors, which revealed the significance of electronegative substitutions on aryl rings (PEOE3) in determining variation of GS inhibitor potency. Furthermore, substitution of acyclic amines with N-substituted cyclic amines appears to be favorable for enhancing GS inhibitor potency by increasing the values of sssN_Cnt and number of aliphatic rings. The models developed are statistically significant and improve our understanding of compounds contributing toward GS inhibitor potency and aid in the rational design of novel potent GS inhibitors.

In silico studies with human DNA topoisomerase-II alpha to unravel the mechanism of in vitro genotoxicity of benzene and its metabolites.

Exposure of humans to benzene present in environment may lead to adverse chronic effects-even at the genetic level. However, the mechanism of its genotoxicity is not well understood. In the present study, in vitro genotoxicity of benzene (BZ) and its major metabolites [p-benzoquinone (BQ), hydroquinone (HQ), catechol (CT), 1,2,4-benzenetriol (BT) and trans-trans muconic acid (MA)] at concentrations 0.5-50 microM, was assessed in Chinese hamster ovary (CHO) cells employing the alkaline Comet assay, cytokinesis blocked micronucleus (CBMN) assay, flow cytometric analysis of micronucleus (flow MN) and chromosome aberration (CA) test. The data revealed significant (P<0.05) concentration-dependent response in all end points. HQ was found to be the most potent DNA damaging metabolite in the Comet assay followed by BQ>BT>CT>BZ>MA. Both CBMN and flow MN assays revealed a good correlation in their results, where BQ and MA exhibited maximum and minimum micronucleus induction respectively. Significant chromosomal aberrations were induced mainly by BQ, BT and HQ, with moderate response shown by CT and BZ and least by MA. The results demonstrated the utility of sensitive techniques like Comet assay and flow cytometry for determination of MN, to quantify in vitro genotoxicity at low levels and also suggested that partly non-repaired DNA damage could cause adverse health effects in human population exposed to benzene. In silico studies using different endpoints of genotoxicity and molecular docking studies with human topoisomerase-II alpha, a major DNA repair enzyme were also conducted. These corroborated the results obtained from the in vitro data, pointing to a direct relationship of the observed genotoxicity with the structural properties and various interactions of metabolites with the enzyme. This comprehensive study demonstrated that genotoxicity of benzene in mammalian cells is mainly due to the inhibition of topoisomerase by the metabolites.

Antiprotozoal Nitazoxanide Derivatives: Synthesis, Bioassays and QSAR Study Combined with Docking for Mechanistic Insight.

In view of the serious health problems concerning infectious diseases in heavily populated areas, we followed the strategy of lead compound diversification to evaluate the near-by chemical space for new organic compounds. To this end, twenty derivatives of nitazoxanide (NTZ) were synthesized and tested for activity against Entamoeba histolytica parasites. To ensure drug-likeliness and activity relatedness of the new compounds, the synthetic work was assisted by a quantitative structure-activity relationships study (QSAR). Many of the inherent downsides - well-known to QSAR practitioners - we circumvented thanks to workarounds which we proposed in prior QSAR publication. To gain further mechanistic insight on a molecular level, ligand-enzyme docking simulations were carried out since NTZ is known to inhibit the protozoal pyruvate ferredoxin oxidoreductase (PFOR) enzyme as its biomolecular target.

A neural network-based QSAR approach for exploration of diverse multi-tyrosine kinase inhibitors and its comparison with a fragment- based approach.

Receptor and non-receptor tyrosine kinases have emerged as clinically useful drug target for treating certain types of cancer. It is well known that tyrosine kinase inhibitors with multi-kinases inhibitory potency are useful in anticancer therapy. In recent study, we have demonstrated application of a novel Group based QSAR (GQSAR) method to assist in lead optimization of multi-tyrosine kinase (PDGFR-beta, FGFR-1 and SRC) inhibitors. Although GQSAR method provides an alternative way to design new compounds, it could not be applied for virtual screening of large databases, because of its limitation to fragment each of the compound in the diverse database. So to circumvent this limitation of GQSAR method, herein we present the development of multi-kinase QSAR model using artificial neural networks. Various simple, easy and fast to calculate 2D/3D descriptors were used in the present analysis. The resulting neural network based QSAR (NN-QSAR) model was found to be statistically significant and provided insight into common structural requirements to inhibit different tyrosine kinases. The NN-QSAR model suggests five descriptors viz. number of rotatable bonds, number of hydrogen bond donors, number of building blocks, polar surface area and sum of nitrogen and oxygen atoms to be of major importance in explaining the activity variation in all the three kinases. In addition, this multi-target QSAR model could be useful to predict the activities of new compounds designed as tyrosine kinase inhibitors.

Application of GQSAR for Scaffold Hopping and Lead Optimization in Multitarget Inhibitors.

Many literature reports suggest that drugs against multiple targets may overcome many limitations of single targets and achieve a more effective and safer control of the disease. However, design of multitarget drugs presents a great challenge. The present study demonstrates application of a novel Group based QSAR (GQSAR) method to assist in lead optimization of multikinase (PDGFR-beta, FGFR-1 and SRC) and scaffold hopping of multiserotonin target (serotonin receptor 1A and serotonin transporter) inhibitors. For GQSAR analysis, a wide variety of structurally diverse multikinase inhibitors (225 molecules) and multiserotonin target inhibitors (162 molecules) were collected from various literature reports. Each molecule in the data set was divided into four fragments (kinase inhibitors) and three fragments (serotonin target inhibitors) and their corresponding two-dimensional fragment descriptors were calculated. The multiresponse regression GQSAR models were developed for both the datasets. The developed GQSAR models were found to be useful for scaffold hopping and lead optimization of multitarget inhibitors. In addition, the developed GQSAR models provide important fragment based features that can form the building blocks to guide combinatorial library design in the search for optimally potent multitarget inhibitors.

A comprehensive structure-activity analysis of protein kinase B-alpha (Akt1) inhibitors.

Protein kinase B (PKB, also known as Akt) belongs to the AGC subfamily of the protein kinase superfamily. Akt1 has been reported as a central player in regulation of metabolism, cell survival, motility, transcription and cell-cycle progression, among the signalling proteins that respond to a large variety of signals. In this study an attempt was made to understand structural requirements for Akt1 inhibition using conventional QSAR, k-nearest neighbour QSAR and novel GQSAR methods. With this intention, a wide variety of structurally diverse Akt1 inhibitors were collected from various literature reports. The conventional QSAR analyses revealed the key role of Baumann's alignment independent topological descriptors along with other descriptors such as the number of hydrogen bond acceptors, hydrogen bond donors, rotatable bonds and aromatic oxygen (SaaOcount) along with molecular branching (chi3Cluster), alkene carbon atom type (SdsCHE-index) in governing activity variation. Further, the GQSAR analyses show that chemical variations like presence of hetero-aromatic ring, flexibility, polar surface area and fragment length present in the hinge binding fragment (in the present case fragment D) are highly influential for achieving highly potent Akt1 inhibitors. In addition, this study resulted in a k-nearest neighbour classification model with three descriptors suggesting the key role of oxygen (SssOE-index) and aromatic carbon (SaaCHE-index and SaasCE-index) atoms electro-topological environment that differentiate molecules binding to Akt1 kinase or PH domain. The developed models are interpretable, with good statistical and predictive significance, and can be used for guiding ligand modification for the development of potential new Akt1 inhibitors.

Characterization of Mixtures. Part 2: QSPR Models for Prediction of Excess Molar Volume and Liquid Density Using Neural Networks.

In our earlier work, we have demonstrated that it is possible to characterize binary mixtures using single component descriptors by applying various mixing rules. We also showed that these methods were successful in building predictive QSPR models to study various mixture properties of interest. Here in, we developed a QSPR model of an excess thermodynamic property of binary mixtures i.e. excess molar volume (V(E) ). In the present study, we use a set of mixture descriptors which we earlier designed to specifically account for intermolecular interactions between the components of a mixture and applied successfully to the prediction of infinite-dilution activity coefficients using neural networks (part 1 of this series). We obtain a significant QSPR model for the prediction of excess molar volume (V(E) ) using consensus neural networks and five mixture descriptors. We find that hydrogen bond and thermodynamic descriptors are the most important in determining excess molar volume (V(E) ), which is in line with the theory of intermolecular forces governing excess mixture properties. The results also suggest that the mixture descriptors utilized herein may be sufficient to model a wide variety of properties of binary and possibly even more complex mixtures.

Rationalizing protein-ligand interactions for PTP1B inhibitors using computational methods.

Protein tyrosine phosphatase 1B inhibitors were reported to have anti-diabetic properties and hence this enzyme has become interesting drug target in the recent time. Huge amount of data is available in public domain about the PTP1B inhibitors in the form of X-ray structures. This study is an attempt to transform this data into useful knowledge which can be directly used to design more effective protein tyrosine phosphatase inhibitors. In this study, we have built quantitative models for activity of co-crystallized protein tyrosine phosphatase inhibitors using two new approaches developed in our group, i.e. receptor-ligand interaction and Structure-based compound optimization, prioritization and evolution based on receptor-ligand interaction descriptors and residue-wise interaction energies as descriptors, respectively. These models have given insights into the receptor-ligand interactions essential for modulating the activity of PTP1B inhibitors. An external validation set of 22 molecules was used to test predictive power of these models on external set molecules.

An improved approximation to the estimation of the critical F values in best subset regression.

Variable selection methods are routinely applied in regression modeling to identify a small number of descriptors which "best" explain the variation in the response variable. Most statistical packages that perform regression have some form of stepping algorithm that can be used in this identification process. Unfortunately, when a subset of p variables measured on a sample of n objects are selected from a set of k (>p) to maximize the squared sample multiple regression coefficient, the significance of the resulting regression is upwardly biased. The extent of this bias is investigated by using Monte Carlo simulation and is presented as an inflation factor which when multiplied by the usual tabulated F ratio gives an estimate of the true 5% critical value. The results show that selection bias can be very high even for moderate-size data sets. Selecting three variables from 50 generated at random with 20 observations will almost certainly provide a significant result if the usual tabulated F values are used. An interpolation formula is provided for the calculation of the inflation factor for different combinations of (n, p, k). Four real data sets are examined to illustrate the effect of correlated descriptor variables on the degree of inflation.

Three-dimensional QSAR using the k-nearest neighbor method and its interpretation.

In this paper we report a novel three-dimensional QSAR approach, kNN-MFA, developed based on principles of the k-nearest neighbor method combined with various variable selection procedures. The kNN-MFA approach was used to generate models for three different data sets and predict the activity of test molecules through each of these models. The three data sets used were the standard steroid benchmark, an antiinflammatory and an anticancerous data set. The study resulted in kNN-MFA models having better statistical parameters than the reported CoMFA models for all the three data sets. It was also found that stochastic methods generate better models resulting in more accurate predictions as compared to stepwise forward selection procedures. Thus, kNN-MFA method represents a good alternative to CoMFA-like methods.

Application of QSPR to mixtures.

In this paper we report an attempt to apply the QSPR approach for the analysis of data for mixtures. This is an extension of the conventional QSPR approach to the analysis of data for single molecules. The QSPR methodology was applied to a data set of experimental measured density of binary liquid mixtures compiled from the literature. The present study is aimed to develop models to predict the "delta" value of a mixture i.e., deviation of the experimental mixture density (MED) from the ideal, mole-weighted calculated mixture density (MCD). The QSPR was investigated in two perspectives (QMD-I and QMD-II) with respect to the creation of training and test sets. The study resulted in significant ensemble neural network and k-nearest neighbor models having statistical parameters r2, q2(10cv) greater than 0.9, and pred_r2 greater than 0.75. The developed models can be used to predict the delta and hence the density of a new mixture. The QSPR analysis shows the importance of hydrogen bond, polar, shape, and thermodynamic descriptors in determining mixture density, thus aiding in the understanding of molecular interactions important in molecular packing in the mixtures.

Modeling and interactions of Aspergillus fumigatus lanosterol 14-alpha demethylase 'A' with azole antifungals.

Recent identification of the sterol 14-alpha demethylase genes (CYP51 A and B) from Aspergillus fumigatus and other species by Mellado et al. (J. Clin. Microbiol. 2001, 39(7), 2431-2438), has opened up possibilities of investigating the interactions of azole antifungals with the enzyme(s) from fungi. This study describes for the first time, a model of the three-dimensional structure of A. fumigatus 14-alpha demethylase (AF-CYP51A), using the crystal structure of Mycobacterium tuberculosis 14-alpha demethylase (PDB code:1EA1) as a template. The paper also describes the various interactions between azole antifungals and the target from A. fumigatus (AF-CYP51A). Quantitative evaluation of these interactions is done using COMBINE analysis to understand contributions of active site residues to ligand activity. It also provides explanation for the activity/inactivity of different ligands for AF-CYP51A.