Selection of 2D/3D molecular descriptors and QSAR modeling of aromatic Morita-Baylis-Hillman adducts with leishmanicidal activities

Aromatic Morita-Baylis-Hillman adducts (MBHA) are a class of compounds which have shown antiparasitic potential in tropical diseases (such as Leishmaniases, Chagas disease, malaria). MBHA analogs have been studied by our research group on synthetical, theoretical, and bioactivity aspects. We present here a variable selection of 2D and 3D molecular descriptors, followed by quantitative structure-activity relationship (QSAR) modeling of thirty-two MBHA bioactive against the parasite Leishmania amazonensis. Descriptors were calculated by E-Dragon online platform. Variable selection was performed using ordered predictor selector (OPS) algorithm, and QSAR models were obtained using partial least squares (PLS). Internal and external validation parameters indicated a robust and predictive model, which may help the synthesis of most potent leishmanicidal compounds.