Prospecção in silico de compostos bioativos contra malária utilizando modelos de previsão de atividade biológica

Abstract

Strategies for drug design aided by computer uses of computational techniques for guidance on the discovery of new drugs. Saving costs and speeding the process of research and development, besides it being able to process a big volume data. The utilization of LQTA-QSAR allows the creation of predicting models of biological activity able to propose novel compounds without the real necessity of in vitro biological activity essays. Malaria is a disease caused by the parasite from genre Plasmodium and affects mainly the Africa and Latin America regions. It is estimated that in 2014 it happened 214 million cases of malaria and 438 thousand cases of death according to World Health Organization. Plasmepsin II is responsible for the metabolism of hemoglobin and it is a valid target for the development of new antimalarial agents. The general objective of this work is to develop a predictive model of biological activity based on the structure of Plasmepsin II inhibitors derivate from allophenylnorstatine. A model was built with 37 samples, Q² = 0.92 and R² = 0.95 with 10 variables and 2 latent variables. The model present good statistical parameters and good predicting ability (Q²ext=0.79). The descriptors could be well correlated to ligand-Plm II interactions. The data obtained serves as guides to the design of new optimized antimalarial agents.