Each drug can potentially be metabolized by different CYP450 isoforms. In the development of new drugs, the prediction of the metabolic fate is important to prevent drug-drug interactions. In the present study, a collection of 580 CYP450 substrates is deeply analyzed by applying multi- and single-label classification strategies, after the computation and selection of suitable molecular descriptors. Cross-training with support vector machine, multilabel k-nearest-neighbor and counterpropagation neural network modeling methods were used in the multilabel approach, which allows one to classify the compounds simultaneously in multiple classes. In the single-label models, automatic variable selection was combined with various cross-validation experiments and modeling techniques. Moreover, the reliability of both multi- and single-label models was assessed by the prediction of an external test set. Finally, the predicted results of the best models were compared to show that, even if the models present similar performances, the multilabel approach more coherently reflects the real metabolism information.

No abstract

No abstract

The integration of ligand- and structure-based strategies might sensitively increase the success of drug discovery process. We have recently described the application of Molecular Electrostatic Potential autocorrelated vectors (autoMEPs) in generating both linear (Partial Least-Square, PLS) and nonlinear (Response Surface Analysis, RSA) 3D-QSAR models to quantitatively predict the binding affinity of human adenosine A3 receptor antagonists. Moreover, we have also reported a novel GPCR modeling approach, called Ligand-Based Homology Modeling (LBHM), as a tool to simulate the conformational changes of the receptor induced by ligand binding. In the present study, the application of both linear and nonlinear 3D-QSAR methods and LBHM computational techniques has been used to depict the hypothetical antagonist binding site of the human adenosine A2A receptor. In particular, a collection of 127 known human A2A antagonists has been utilized to derive two 3D-QSAR models (autoMEPs/PLS&RSA). In parallel, using a rhodopsin-driven homology modeling approach, we have built a model of the human adenosine A2A receptor. Finally, 3D-QSAR and LBHM strategies have been utilized to predict the binding affinity of five new human A2A pyrazolo-triazolo-pyrimidine antagonists finding a good agreement between the theoretical and the experimental predictions.

No abstract

Nowadays, in medicinal chemistry adenosine receptors represent some of the most studied targets, and there is growing interest on the different adenosine receptor (AR) subtypes. The AR subtypes selectivity is highly desired in the development of potent ligands to achieve the therapeutic success. So far, very few ligand-based strategies have been investigated to predict the receptor subtypes selectivity. In the present study, we have carried out a novel application of the multilabel classification approach by combining our recently reported autocorrelated molecular descriptors encoding for the molecular electrostatic potential (autoMEP) with support vector machines (SVMs). Three valuable models, based on decreasing thresholds of potency, have been generated as in series quantitative sieves for the simultaneous prediction of the hA(1)R, hA(2A)R, hA(2B)R, and hA(3)R subtypes potency profile and selectivity of a large collection, more than 500, of known inverse agonists such as xanthine, pyrazolo-triazolo-pyrimidine, and triazolo-pyrimidine analogues. The robustness and reliability of our multilabel classification models were assessed by predicting an internal test set. Finally, we have applied our strategy to 13 newly synthesized pyrazolo-triazolo-pyrimidine derivatives inferring their full adenosine receptor potency spectrum and hAR subtypes selectivity profile.

Quantitative structure-activity relationship (QSAR) analysis has been frequently utilized as a computational tool for the prediction of several eco-toxicological parameters including the acute aquatic toxicity. In the present study, we describe a novel integrated strategy to describe the acute aquatic toxicity through the combination of both toxicokinetic and toxicodynamic behaviors of chemicals. In particular, a robust classification model (TOXclass) has been derived by combining Support Vector Machine (SVM) analysis with three classes of toxicokinetic-like molecular descriptors: the autocorrelation molecular electrostatic potential (autoMEP) vectors, Sterimol topological descriptors and logP(o/w) property values. TOXclass model is able to assign chemicals to different levels of acute aquatic toxicity, providing an appropriate answer to the new regulatory requirements. Moreover, we have extended the above mentioned toxicokinetic-like descriptor set with a more toxicodynamic-like descriptors, as for example HOMO and LUMO energies, to generate a valuable SVM classifier (MOAclass) for the prediction of the mode of action (MOA) of toxic chemicals. As preliminary validation of our approach, the toxicokinetic (TOXclass) and the toxicodynamic (MOAclass) models have been applied in series to inspect both aquatic toxicity hazard and mode of action of 296 chemical substances with unknown or uncertain toxicodynamic information to assess the potential ecological risk and the toxic mechanism.

In recent years, more and more attention has been paid on learning in structured domains, e.g. Chemistry. Both Neural Networks and Kernel Methods for structured data have been proposed. Here, we show that a recently developed technique for structured domains, i.e. PCA for structures, permits to generate representations of graphs (specif- ically, molecular graphs) which are quite effective when used for predic- tion tasks (QSAR studies). The advantage of these representations is that they can be generated automatically and exploited by any tradi- tional predictor (e.g., Support Vector Regression with linear kernel)