Optimization of Gaussian Kernel Function in Support Vector Machine aided QSAR studies of C-aryl glucoside SGLT2 inhibitors

Abstract

The present investigations include utility of latest statistical algorithm Support Vector Machine (SVM) to identify non-linear structure activity relationship between IC50 values and structures of C-aryl glucoside SGLT2 inhibitors. Training dataset consisted of forty molecules and the remaining six molecules were chosen for test set validation. SVM under Gaussian Kernel Function yielded non-linear QSAR models. Forward selection algorithm was applied after pruning and redundancy check on molecular descriptors. Internal validations of QSAR models have been achieved using R CV (2) (LOO), PRESS, SDEP and Y-Scrambling. SVM aided non-linear models are more efficient when optimization of Gaussian Kernel Function was introduced. Non-linear QSAR studies further identified atomic van der Waals volumes, atomic masses, sum of geometrical distances between O..S and degree of unsaturation as molecular descriptors and crucial structural requirements to model IC50 of C-aryl glucoside derivatives.