Abstract
Certain of the fundamental concepts underlying the utilization of comparative molecular moment (CoMMA) descriptors as measures of three-dimensional molecular similarity are reviewed. The results of a principal component regression (PCR) analyses of the five data sets previously examined by partial least squares (PLS) calculations are provided. The results further substantiate the utility of the CoMMA descriptors in predicting chemical and biological activity.
Keywords: Molecular Moments, Computer Aided Drug Discovery, CoMMA descriptors, Cartesian reference frame, quadrupolar tensor, MATLAB Statistics Toolbox, QSAR regressions