Identification of Natural Products with Potential Activity against Leishmania amazonensis using computational models and experimental corroboration

Leishmaniasis is one of the most important neglected tropical diseases according to the World Health Organization. The available drugs are expensive, not sufficiently effective, have serious cytotoxic effects and parasitic resistance has increased in the last years. In the present work, a virtual screening protocol was used to identify new natural compounds potentially active against Leishmania spp. using machine learning-based models. Three vegetable origin compounds were selected by using a multiclassifier composed by models developed with k-nearest neighbor, classification tree, Multilayer perceptron and Support Vector Machine; all these models for Leishmania amazonensis promastigote form were developed with WEKA software. The selected compounds showed in vitro activity against L. amazonensis (MHOM/BR/77/LTB0016) promastigotes with CI₅₀ lower than 1 µg/mL using 96-well plates and resazurine fluorescence method.