QSAR on aryl-piperazine derivatives with activity on malaria

In this work we offer linear regression models on a set of aryl-piperazine derivatives that are obtained by exploring a pool containing 1497 Dragon molecular descriptors, in order to establish the best relationships linking the molecular structure characteristics to their exhibited potencies against chloroquine resistant and chloroquine sensitive strains of Plasmodium falciparum parasite. The adjustment of the training molecular set together with the performance achieved during the internal and external validation processes leads to predictive QSAR models. In addition, we derive alternative linear models based on the Coral methodology, which lead to satisfactory results. We apply the final equations to predict the activity on some unknown compounds having non-observed activities.

► Linear QSAR for predicting the antiplasmodial activity of aryl-piperazine derivatives. ► Molecular structure translated into Dragon, Recon, and Coral molecular descriptors. ► Geometry optimization with Semiempirical Methods. ► Validation of the QSAR models with external test set and Cross-Validation. ► Prediction of structures with unknown antimalarial activities.