QSAR models for the prediction of anti-HIV activity using genetic algorithm multiple linear regression (GA-MLR) and back propagation artificial neural networks (BP-ANN) analysis

The study of Quantitative Structure-Activity Relationships (QSARS) was conducted to explore the correlation between a set of ۱۰۷ derivatives of ۱-[۲-hydroxyethoxy-methyl]-۶-(phenylthio) thymine (HEPT) as non-nucleoside reverse transcriptase inhibitors, which exhibit notable anti-HIV-۱ activity, and their molecular descriptors. This was achieved using genetic algorithm-based multiple linear regression (GA-MLR) and back propagation artificial neural networks (BP-ANN) techniques. The chemical structures of the derivatives were created using Gauss View ۰۵ software, and the structures were further optimized using Gaussian ۰۹ software with the DFT-B۳LYP functional and ۶-۳۱G* basis sets. Results showed that the octanol/water partition coefficient is the main structural features contributing to the anti-HIV-۱ activity of the HEPT derivatives.

کلیدواژه ها:

QSARS ، Anti-HIV-۱ activity ، GA-MLR ، Gauss View ۰۵ software