Predição da estrutura tridimensional de proteínas utilizando o método CReF com informações de contato

• Autor(a) principal: Marques, Flavielle Blanco
• Data de Publicação: 2021
• Tipo de documento: Dissertação
• Idioma: por
• Título da fonte: Biblioteca Digital de Teses e Dissertações da PUC_RS
• Texto Completo: http://tede2.pucrs.br/tede2/handle/tede/9787
• topic: Bioinformática; Predição de Estruturas de Proteínas; Contato; Bioinformatics; Protein Structure Prediction; Contact; CIENCIA DA COMPUTACAO::TEORIA DA COMPUTACAO

Protein folding is a problem with a quick solution in nature. However, look for all conformations that a sequence of amino acid residues can assume is not a trivial task. In this way, several predictors are trying to create solutions for protein structure prediction. One of these is the CReF (Central Residue Fragment-based method). In the last years, approaches using the inter-residue contact information had promissory results to protein structure prediction. Thus, we incorporated contact information into the CReF method. However, functionals and methodological changes were necessary. Questions about conformation evaluation and sampling were central to contact information uses were effective. In the first, we created na RMSDpredict model. It uses distance-dependent atomic potential and short, medium, and long-range inter-residue contacts. In the second, a molecular simulation module based on simulated annealing was coupling. Additionally, we proposed a function based on the same terms in the model. It behaves such as function energy. Furthermore, we simulated and evaluated a protein sample. If the inter-residue contact information is available, it is possible to find protein conformations with low energy and RMSD. Hence, we created a CReF new version with contact information, a function to calculate the energy system, and a simulation method.