Application of molecular topology to the prediction of mosquito repellents of a group of terpenoid compounds.

A topological-mathematical model based on multilinear regression analysis has been built to search new terpenoid actives as mosquito repellents. The structural depiction was performed using topological indices, and a four-variable model for the prediction of corrected repellent ratio (r (2) = 0.9672, Q (2) = 0.9435) was selected. The model was checked by cross-validation, internal validation, and randomization test. The results confirm its capability to predict the property analyzed. After carrying out a virtual screening upon such a model, new structures with potential repellent activity against mosquitoes are proposed.