TOPS-MODE based QSARs derived from heterogeneous series of compounds. Applications to the design of new anti-inflammatory compounds

A new application of TOPological Sub-structural MOlecular DEsign (TOPS-MODE) was carried out in anti-inflammatory compounds using computer-aided molecular design. Two series of compounds, one containing anti-inflammatory and the other containing nonanti-inflammatory compounds were processed by a k-means cluster analysis in order to design the training and prediction sets. A linear classification function to discriminate the anti-inflammatory from the inactive compounds was developed. The model correctly and clearly classified 88% of active and 91% of inactive compounds in the training set. More specifically, the model showed a good global classification of 90%, that is, (399 cases out of 441). While in the prediction set, they showed an overall predictability of 88% and 84% for active and inactive compounds, being the global percentage of good classification of 85%. Furthermore this paper describes a fragment analysis in order to determine the contribution of several fragments towards anti-inflammatory property, also the present of halogens in the selected fragments were analyzed. It seems that the present TOPS-MODE based QSAR is the first alternate general 'in silico' technique to experimentation in anti-inflammatory discovery.     

Computer-aided rodent carcinogenicity prediction

The potential of the computer program PASS (Prediction Activity Spectra for Substances) to predict rodent carcinogenicity for chemical compounds was studied. PASS predicts carcinogenicity of chemical compounds on the basis of their structural formula and of structure-activity relationship analysis of known carcinogens and non-carcinogens. The data on structures and experimental results of 2-year carcinogenicity assays for 412 chemicals from the NTP (National Toxicological Program) and 1190 chemicals from the CPDB (Carcinogenic Potency Database) were used in our study. The predictions take into consideration information about species and sex of animals. For evaluation of the predictive accuracy we used two procedures: leave-one-out cross-validation (LOO CV) and leave-20%-out cross-validation. In the last case we randomly divided the studied data set 20 times into two subsets. The data from the first subset, containing 80% of the compounds, were added to the PASS training set (which includes about 46,000 compounds with about 1500 biological activity types collected during the last 20 years to predict biological activity spectra), the second subset with 20% of the compounds was used as an evaluation set. The mean accuracy of prediction calculated by LOO CV is about 73% for NTP compounds in the 'equivocal' category of carcinogenic activity and 80% for NTP compounds in the 'evidence' category of carcinogenicity. The mean accuracy of prediction for the CPDB database is 89.9% calculated by LOO CV and 63.4% calculated by leave-20%-out cross-validation. Influence of incorporation of species and sex data on the accuracy of carcinogenicity prediction was also investigated. It was shown that the accuracy was increased only for data on male animals.     

Three-dimensional quantitative structure activity relationships (3-D-QSAR) of antihyperglycemic agents.

A three-dimensional quantitative structure activity relationship study (3-D-QSAR) was performed on a set of thiazolidinedione antihyperglycemic agents using the comparative molecular field analysis (CoMFA) method. The CoMFA models were derived from a training set of 53 compounds. Fifteen compounds, which were not used in model generation were used to validate the CoMFA models. All the compounds were superimposed to the template structure by atom-based and shape-based strategies. The SYBYL QSAR rigid body field fit was also used for aligning the ligands. A total of twelve different alignments were generated. The resulting models exhibited a good cross-validated r2cv values (0.624-0.764) and the conventional r2 values (0.689-0.921). A more robust cross-validation test using cross-validation by 2 groups (leave half out method) was performed 100 times to ascertain the predictiveness of the CoMFA models. The mean of r2cv values from 100 runs ranged from 0.611-0.690. Few models exhibited good external predictivity. These models were then used to define a hypothetical receptor model for antihyperglycemic agents.     

The systematic modeling studies and free energy calculations of the phenazine compounds as anti-tuberculosis agents

Abstract

Phenazine compounds have good activity against Mycobacterium tuberculosis (MTB). Based on the reported activities that were obtained in MTB H37Rv, a three-dimensional quantitative structure–activity relationship (3D-QSAR) model was built to design novel compounds against MTB. A fivefold cross-validation method and external validation were used to analyze the accuracy of forecasting. The model has a cross-validation coefficient q²=0.7 and a non-cross-validation coefficient r² = 0.903, indicating that the model has good predictive possibility. The design of anti-pneumococcus MTB compounds was guided by the obtained 3D-QSAR model, and several compounds with better activity were obtained. To test the activity of these compounds, molecular docking, molecular dynamics simulation, and post-simulation analysis of the already reported drug targets in MTB were carried out. Among the total 15 drug targets, only three targets (Rv2361c, Rv2965c, and Rv3048c) were selected based on the docking results. Initial results reported that these compounds possessed good inhibition activity for Rv2361c. The top nine complexes of Rv2361 ligands were only subjected to MD simulation which resulted in a stable dynamics of the structures and showed a residual fluctuation in inhibitors binding pocket. Free energy reported that overall, the derivatives hold strong energy against the protein target. Energetic contribution results showed that residues, Asp76, Arg80, Asn124, Arg127, Arg244, and Arg250, play a major role in total energy. Systems biology approach validates shortlisted drug effect on the entire system which might be useful to predict potential drug in wet lab as well.

Communicated by Ramaswamy H. Sarma     

Cross-Platform Toxicogenomics for the Prediction of Non-Genotoxic Hepatocarcinogenesis in Rat

In the area of omics profiling in toxicology, i.e. toxicogenomics, characteristic molecular profiles have previously been incorporated into prediction models for early assessment of a carcinogenic potential and mechanism-based classification of compounds. Traditionally, the biomarker signatures used for model construction were derived from individual high-throughput techniques, such as microarrays designed for monitoring global mRNA expression. In this study, we built predictive models by integrating omics data across complementary microarray platforms and introduced new concepts for modeling of pathway alterations and molecular interactions between multiple biological layers. We trained and evaluated diverse machine learning-based models, differing in the incorporated features and learning algorithms on a cross-omics dataset encompassing mRNA, miRNA, and protein expression profiles obtained from rat liver samples treated with a heterogeneous set of substances. Most of these compounds could be unambiguously classified as genotoxic carcinogens, non-genotoxic carcinogens, or non-hepatocarcinogens based on evidence from published studies. Since mixed characteristics were reported for the compounds Cyproterone acetate, Thioacetamide, and Wy-14643, we reclassified these compounds as either genotoxic or non-genotoxic carcinogens based on their molecular profiles. Evaluating our toxicogenomics models in a repeated external cross-validation procedure, we demonstrated that the prediction accuracy of our models could be increased by joining the biomarker signatures across multiple biological layers and by adding complex features derived from cross-platform integration of the omics data. Furthermore, we found that adding these features resulted in a better separation of the compound classes and a more confident reclassification of the three undefined compounds as non-genotoxic carcinogens.     

Megavariate data analysis of mass spectrometric proteomics data using latent variable projection method

There are many data mining techniques for processing and general learning of multivariate data. However, we believe the wavelet transformation and latent variable projection method are particularly useful for spectroscopic and chromatographic data. Projection based methods are designed to handle hugely multivariate nature of such data effectively. For the actual analysis of the data we have used latent variable projection methods such as principal component analysis (PCA) and partial least squares projection to latent structures based discriminant analysis (PLS-DA) to analyze the raw data presented to the participants of the First Duke Proteomics Data Mining Conference. PCA was used to solve problem #1 (clustering problem) and the PLS-DA was used to solve problem #2 (classification problem). The idea of internal and external cross-validation was used to validate the model obtained from the classification analysis. The simple two-component PLS-DA model obtained from the analysis performed well. The model has completely separated the two groups from all the data. The same model applied on two-thirds of the data showed good performance by external validation with independent test set of remaining 13 specimens obtained by setting aside the spectra of every third specimen (accuracy of 85%).     

Binding conformations and QSAR of CA-4 analogs as tubulin inhibitors

A theoretical study on the binding conformations and the quantitative structure–activity relationship (QSAR) of combretastatin A4 (CA-4) analogs as inhibitors toward tubulin has been carried out using docking analysis and comparative molecular field analysis (CoMFA). The appropriate binding orientations and conformations of these compounds interacting with tubulin were revealed by the docking study; and a 3D-QSAR model showing significant statistical quality and satisfactory predictive ability was established, in which the correlation coefficient (R²) and cross-validation coefficient (q²) were 0.955 and 0.66, respectively. The same model was further applied to predict the pIC₅₀ values for 16 congeneric compounds as external test set, and the predictive correlation coefficient R²_pred reached 0.883. Other tests on additional validations further confirmed the satisfactory predictive power of the model. In this work, it was very interesting to find that the 3D topology structure of the active site of tubulin from the docking analysis was in good agreement with the 3D-QSAR model from CoMFA for this series of compounds. Some key structural factors of the compounds responsible for cytotoxicity were reasonably presented. These theoretical results can offer useful references for understanding the action mechanism and directing the molecular design of this kind of inhibitor with improved activity.     

A novel model to predict O-glycosylation sites using a highly unbalanced dataset

In silico approaches have become an alternative method to study O-glycosylation. In this paper, we developed a linear interpretable model for O-glycosylation prediction based on an unbalanced dataset, analyzing the underlying biological knowledge of glycosylation. A training set of 4446 sites involving 468 positive sites and 3978 negative sites was developed during this research. The sites were encoded using the amino acid index (AAindex), and the forward stepwise procedure utilized for feature selection. The linear discriminant analysis with an equal a priori probability (PP-LDA) was employed to develop the interpretable model. Performance of the model was verified using both the internal leave-one-out cross-validation and external validation methods. Two non-linear algorithms, the supervised support vector machine and the unsupervised self-organizing competitive neural network, were used as comparisons. The PP-LDA model exhibited improved classification results with accuracy of 82.1?% for cross-validations and 80.3?% for external prediction. Further analysis of this linear model indicated that the properties at position R(1) and the properties relative to hydrophobicity contributed more to the glycosylation prediction. However, the alpha and turn propensities at the C-terminal, together with physicochemical properties at the N-terminal, are also relative to the glycosylation activity. This model is not only capable of predicting the possibility of glycosylation using an unbalanced dataset, but is also helpful to understand the underlying biological mechanisms of glycosylation. Considering the publicly accessibility of our prediction model, a downloadable program is provided in our supply materials.     

Comparative molecular field analysis (CoMFA) of phthalazine derivatives as phosphodiesterase IV inhibitors

A comparative molecular field analysis (CoMFA) of phthalazine class of phosphodiesterase IV (PDE IV) inhibitors has been performed to correlate their chemical structures with their observed biological activity. A statistically valid model with good correlative and predictive power is reported. The leave one out cross-validation study gave cross-validation r(2)(cv) of value 0.507 at six optimum components and conventional r(2) of value 0.98. The predictive ability of the model was tested by predicting the seven molecules belonging to the test set giving predictive correlation coefficient of 0.59. This model is potentially helpful in the design of novel and more potent PDE IV inhibitors.