2D-autocorrelation descriptors for predicting cytotoxicity of naphthoquinone ester derivatives against oral human epidermoid carcinoma.

A QSAR study was developed, employing 2D-autocorrelation descriptors and a set of 37 naphthoquinone ester derivatives, in order to model the cytotoxicity of these compounds against oral human epidermoid carcinoma (KB). A comparison with other approaches such as the BCUT, Galvez topological charge indexes, Randić molecular profile, Geometrical, and RDF descriptors was carried out. Mathematical models were obtained by means of the multiple regression analysis (MRA) and the variables were selected using genetic algorithm. Based on the statistical results the 2D-autocorrelation descriptors were considered the best and were able to describe more than 84.2% of the variance in the experimental activity once we controlled for outliers.