Protein bioinformatics: overview and computational strategies to determine protein interaction patterns to assist in disease control

• Ano de defesa: 2022
• Autor(a) principal: Gomes, Isabela de Souza
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: eng
• Instituição de defesa: Universidade Federal de Viçosa; Ciência da Computação
• Palavras-chave em Português: Bioinformática; Interação proteína-proteína - Processamento de dados; Peptídeos; Aprendizado do computador; Dinâmica molecular; Proteínas; Ciência da Computação
• Link de acesso: https://locus.ufv.br//handle/123456789/30488; https://doi.org/10.47328/ufvbbt.2022.550

Proteins are fundamental biomolecules for the metabolism of living beings and have several biotechnological uses. The computational study of this class of macromolecules allows the ex- pansion of knowledge and speed in research applications, such as catalytic processes, three- dimensional conformations, enzyme inhibition, molecular engineering, among others. In this dissertation, we present a set of papers that purposes computational strategies to study three- dimensional structures of proteins. In the first work, we combine an in-house developed machine learning strategy with docking, MM-PBSA, and metadynamics simulations to detect potential inhibitors for SARS-COV-2 main protease. Computational strategies can help to speed up the process of drug discovery, reducing the time and cost of wet-lab experiments because they will be focused on fewer molecules. Our work points out six ligands that have a good interaction with our target in its active pocket, indicating an inhibitor behavior. We highlighted the strongest interaction of our experiments, M pro -mirabegron complex, which was used as input for subse- quent in vitro assays to validate the inhibition potential suggested by in silico experiments. In the second paper, we present a literature review of several bioinformatics tools for the study of proteins. The article is a very detailed material to support the choice of students and profession- als for the most appropriate tool for a particular application. In the third work, we introduced the Propedia database for protein-peptide identification, which comprises over 19,000 high- resolution structures from the Protein Data Bank. Protein-peptide interactions can be useful for predicting, classifying, and scoring complexes or for designing new molecules. The main ad- vantage of Propedia over other peptide databases is that it allows a more comprehensive analysis of similarity and redundancy. The papers presented here provide an overview of the diversity of protein bioinformatics study and some of its applications in biological problems. Keywords: Bioinformatics. Protein-protein interaction. Peptide-protein interaction. Machine learning. Docking. Molecular dynamics. Proteins.