Modelagem de redes de regulação celular aplicada ao câncer de mama

• Ano de defesa: 2019
• Autor(a) principal: Ferreira, Luiz Henrique Oliveira
• Orientador(a): Castro, Maria Clícia Stelling de
• Banca de defesa: Carels, Nicolas , Silva, Dilson
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade do Estado do Rio de Janeiro
• Programa de Pós-Graduação: Programa de Pós-Graduação em Ciências Computacionais
• Departamento: Centro de Tecnologia e Ciências::Instituto de Matemática e Estatística
• País: BR
• Palavras-chave em Português: Redes booleanas; Sistemas complexos; Modelagem de redes de regulação gênica; Bioinfomática; Modelagem de dados; Algoritmos computacionais
• Palavras-chave em Inglês: Boolean Networks; Complex Systems; Modeling of Gene Regulatory Networks
• Área do conhecimento CNPq: CNPQ::CIENCIAS EXATAS E DA TERRA::CIENCIA DA COMPUTACAO
• Link de acesso: http://www.bdtd.uerj.br/handle/1/7672

Gene Regulatory Networks (GRNs) describe gene interactions that eventually control cell processes like growth and dissemination involved in the hallmarks of cancer. Hallmarks have contributed to consider cancer by several authors as a network disease. The literature describes several models used to help biologists to make predictions, prognosis and even specialized treatment when some disturbances of the GRN lead to a sick condition. Among those, Boolean networks emerge as a simple, yet accurate, tool for modeling GRNs. Nevertheless, when it comes to inferring Boolean functions from gene expression in large networks, it is advisable to establish some form of network simplification otherwise the problem may become practically intractable. The method describes herein takes into account information about the interactome to build a network where each node represents a gene described by the degree entropy of its protein as a key to reduce the size of the network. This process enables to focus only on genes encoding the main hubs; the ones with the largest potential of interfering on the overall network behavior. This work describes an objective algorithm for selecting gene targets as building blocks in GRN modeling. That algorithm allows broader experimentation with GRN models, especially when extensive biological knowledge is not available for screening the genes suitable for modeling. Simulations made with a selected model showed similar behavior to what is supposed to be expected from real cells such as attractor robustness (steady states), for instance. Future works may explore additional criteria for network simplification based on alternative but relevant information, such as expression levels as well as network control