QSPR Models to Predict Thermodynamic Properties of Alkenes Using Genetic Algorithm and Backward- Multiple Linear Regressions Methods

Abstract

Quantitative structure–property relationship (QSPR) models establish relationships between different types of structural information to their properties. In the present study the relationship between the molecular descriptors and quantum properties consist of the heat capacity (Cv/J mol-1K-1) entropy (S/J mol-1K-1) and thermal energy (Eth/kJ mol-1) of 100 alkenes is represented. Genetic algorithm (GA) and backward-multiple linear regressions (BW-MLR) were successfully developed to predict quantum properties of alkenes. Molecular descriptors were calculated with Dragon software and the genetic algorithm (GA) method was used to selected important molecular descriptors. The quantum properties were obtained from quantum-chemistry technique at the Hartree-Fock (HF) level using the ab initio 6-31G* basis sets. The predictive powers of the BW-MLR models were discussed by using leave-one-out (LOO) cross-validation and external test set. Results showed that the predictive ability of the models was satisfactory, and the 2D matrix-based descriptors, topological, edge adjacency and Connectivity indices could be used to predict the mentioned properties of 100 alkenes