QSAR Study of Capacity Factors by Quantum Chemical Descriptors and Using PLS and LS-SVM Methods
Saeed Jameh Bozorghi, Али Ниази · 2010
Introduction: quantitative structure-activity relationship (QSAR) study is one of best chemometrics methods for prediction of chemical and biological properties of various compounds. Aim: A quantitative structure-activity relationship (QSAR) study is suggested for the prediction of capacity factors (log k) of 25 substances as solutes to two different stationary phases (polyethylene–silica and polyethylene–alumina) were analyzed to their quantum chemical descriptors and related to their retention behavior as expressed by the logarithms of their capacity factors (log k). Material and Method: Ab initio theory was used to calculate some quantum chemical descriptors including electrostatic potentials and local charges at each atom, HOMO and LUMO energies, etc. Modeling of the log k as a function of molecular structures was established by means of the partial least squares (PLS) and least squares support vector machines (LS-SVM). These models were applied for the prediction of the capacity factors, which were not in the modeling procedure. Results: The resulted models showed high prediction ability with root mean square error of prediction of 6.5621 and 0.4960 for PLS and LS-SVM, respectively. Conclusion: this research showed that LS-SVM method has a very good ability for prediction of capacity factors (log k) for benzene derivatives.