No abstract

In this study, we present a new model that has been developed for the prediction of θ (lower critical solution temperature) using a database of 169 data points that include 12 polymers and 67 solvents. For the characterization of polymer and solvent molecules, a number of molecular descriptors (topological, physicochemical,steric and electronic) were examined. The best subset of descriptors was selected using the elimination selection-stepwise regression method. Multiple linear regression (MLR) served as the statistical tool to explore the potential correlation among the molecular descriptors and the experimental data. The prediction accuracy of the MLR model was tested using the leave-one-out cross validation procedure, validation through an external test set and the Y-randomization evaluation technique. The domain of applicability was finally determined to identify the reliable predictions.

No abstract

This paper presents the results of an optimization study on biaryl piperidine and 4-amino-2-biarylurea MCH1 receptor antagonists, which was accomplished by using quantitative-structure activity relationships (QSARs), classification and virtual screening techniques. First, a linear QSAR model was developed using Multiple Linear Regression (MLR) Analysis, while the Elimination Selection-Stepwise Regression (ES-SWR) method was adopted for selecting the most suitable input variables. The predictive activity of the model was evaluated using an external validation set and the Y-randomization technique. Based on the selected descriptors, the Support Vector Machines (SVM) classification technique was utilized to classify data into two categories: "actives" or "non-actives". Several attempts were made to optimize the scaffold of most potent compounds by inducing various structural modifications. Potential derivatives with improved activities were identified, as they were classified "actives" by the SVM classifier. Their activities were estimated using the produced MLR model. A detailed analysis on the model applicability domain defined the compounds, whose estimations can be accepted with confidence.

A novel QSPR model is developed and evaluated for the prediction of diamagnetic susceptibility. The model was produced using the Multiple Linear Regression (MLR) technique on a database that consists of 406 organic compounds involving a diverse set of chemical structures. The accuracy of the QSPR model (R 2 ¼ 0.88) is illustrated using various evaluation techniques, such as leave-one-out procedure (Q 2 ¼ 0.87) and validation through an external test set (R 2 pred ¼ 0.89). The study leads to the conclusion that three physical -topological descriptors affect significantly the diamagnetic susceptibility: Polar Surface Area (PSAr), Principal Moment of Inertia X (PMIX), and Diameter (Diam).