Gapin: uma ferramenta para visualização e análise de redes de interações atômicas intermoleculares

• Ano de defesa: 2019
• Autor(a) principal: Biharck Muniz Araújo
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Minas Gerais; Brasil; Programa de Pós-Graduação em Bioinformatica; UFMG
• Palavras-chave em Português: Protein-Protein-Interface; Macromolecules; PPI; Bioinformática; Domínios e Motivos de Interação entre Proteínas; Modelos Moleculares; Desenvolvimento de Medicamentos
• Link de acesso: http://hdl.handle.net/1843/77747

Aiming to extract useful knowledge mainly through representations of the interfaces in networks of contacts, several computational methods, databases, and innovative tools have been proposed for visualization and analysis of data of biomolecular interactions. However, such initiatives tend to be specialized in a particular type of interface, (such as protein-protein or protein-ligand, for instance), and do not easily generalize to any kind of interfaces, regardless of the biomolecules involved. And it is more than that: What is a good way to visualize such interactions networks, focusing on the design of drug candidates? In order to add efforts to these challenges, we propose a tool called GAPIN - Grouped and Aligned Protein Interface Networks, which is a 100 % web application, with high availability, portability, usability and convenience that modern browsers can offer. The main data input for GAPIN is PDB files. GAPIN will define interfaces between biomolecules as graphs at the atomic level. A granularity at this level allows independent visualizations of the involved biomolecules, whether among proteins, nucleic acids, carbohydrates, lipids, ligands, ions or even waters. GAPIN is able to contrast the PDB structures rendered with its respective graphs, in two levels of granularity: the first one, with nodes representing atoms of the interface; the second, with nodes representing communities of atoms (the result of graph clustering) according to the density of the edges, forming modular or higher-level graphs. GAPIN also provides an option for alignment and similarity of modularized graphs. We show in this work that high-level graphs can help in the identification and characterization of Spots, regions at the protein-protein interfaces that group together residues with a relevant contribution to the ⌧ of binding. Literature shows strong evidence that such regions tend to be potential targets for drug candidates. Aiming to highlight overal spectrum of GAPIN, two experiments were developed: the first one involving the alignment of hydrophobic regions in a database of different serine-peptidases and inhibitors; the other one, involving the characterization of Spots in an alanine mutagenesis data set. We have shown that GAPIN was able to identify, by alignment, non-trivial hydrophobic patterns in the first database. It was also able to reveal, in the second database, a curious correlation between the contact areas of nodes in modularized graphs and the presence of energetic Spots. The outcome of this work expects the possibility to demonstrate the visual versatility and analytical potentiality of the GAPIN tool for the study of a huge variety of intermolecular interfaces, with effective power to help researchers from different specialties to get a better understanding about the topological and physicochemical properties of potential therapeutic targets for innovative drugs.