Deep learning for prediction and structural characterization of glycosylated toxins from Bothrops snakes

Abstract

Protein glycosylation is a co-translational and post-translational modification that involves the insertion of carbohydrates into nascent polypeptides. The presence of glycans on proteins has an impact on several physiological events. In snake toxins, glycosylation is pointed out as an important structural and functional characteristic. The fusion of analytical and computational methods constitutes a promising approach for studying the role of glycosylation in the structure and biological activity of proteins. In this sense, the present project aims to extend the identification and structural characterization carried out in the glycoproteome of botropic venoms and develop a computational approach for predicting the structure of glycoproteins, using a deep learning model. In the first year of the Direct Doctorate, the characterization of the machinery involved in the N-glycosylation of toxins was carried out through the integration of multiomics data from Bothrops jararaca. This study made it possible to investigate the molecular bases involved in the synthesis and processing of toxins and the evolutionary aspects of N-glycosylation. Furthermore, this project will allow the identification of important sites for the interaction of toxins with their biological targets and will assist in the discovery of molecules with therapeutic potential. (AU)