QSPR Based on Genetic Algorithm- Multiple Linear Regression and PLS Methods for Prediction of Physico-Chemical Properties of Sulfonamide Drugs

Quantitative structure- property relationships (QSPR) modeling are important tools in drug study. In this study, the QSPR models have been developed for a set of sulfa drug compounds. The logarithm of the octanol/water partition coefficient (logP), and melting point (mp) of these compounds were studied by genetic algorithm based multiple linear regression (GA-MLR) and partial least squares (PLS) methods. Comparison of the quality of the MLR and PLS models shows that the PLS models have substantially better predictive capabilities than the MLR models (higher R2 values , lower RMSE values).