Predictive QSAR study of chalcone derivatives cytotoxicity activity against HT-29 human colon adenocarcinoma cell lines

•A QSAR model of the cytotoxicity against HT-29 cell lines by chalcones was developed.•The best model contained 7 descriptors selected from 1433 by ERM.•The validation procedures revealed very good predictive attributes of the model.•The model implied a high dependence of the activity on the number of 10-member rings.

Chalcones and their derivatives exhibit a wide range of important pharmacological activities; among the most relevant ones is their anticancer activity. For this reason we performed a predictive Quantitative Structure–Activity Relationships (QSAR) analysis of the anticancer activity against HT-29 human colon adenocarcinoma cell lines by chalcones. The best linear model constructed from 136 molecular structures incorporated seven molecular descriptors, selected from more than a thousand geometrical, topological, quantum-mechanical and electronic types of descriptors, showed good predictive ability. The model analysis showed that the descriptors with highest importance on the activity are the number of 10-member rings and a BCUT descriptor weighted by the van der Waals volume.