Computational methods in Biology: cancer biomarkers, protein networks and lateral gene transfer

• Autor(a) principal: Heberle, Henry
• Data de Publicação: 2019
• Tipo de documento: Tese
• Idioma: eng
• Título da fonte: Biblioteca Digital de Teses e Dissertações da USP
• Texto Completo: http://www.teses.usp.br/teses/disponiveis/55/55134/tde-15102019-145225/
• topic: Biological network Visualization; Biomarcadores de câncer; Cancer biomarker; Lateral gene transfer; Priorização de proteínas, Visualização de redes biológicas; Protein prioritization; Transmição Lateral de gene; Tree visualization; Visualização de árvores

Molecular Biology is a branch within Science of great importance. Despite the fact it studies microscopic entities, the volume and complexity of information are great. The applications are varied and can be of global interest, such as the spread of antibiotic resistance genes among bacteria and new methods for diagnostic and prognostic of cancer. By understanding biomolecular mechanisms, scientists can define treatments for diseases, support the decisions made by patients, identify the influence of intestinal microbiota over physical and psychological conditions, find cause and source of microbial antibiotic resistance, among many other applications. Computer Science plays key roles in this context, such as enabling complex data analyzes by specialists, creating models that simulate biological structures and processes, and by providing algorithms for extracting information encoded in biological data. During my doctorate, we explored those mechanisms in three main levels: quantification of proteins from cells, analysis of interactions that happen inside cells, and the comparison of genomes and their genetic history. This manuscript reports different projects, four of them already published in scientific journals. They comprise the discovery of candidate proteins for cancer biomarkers, the visual analysis of protein-protein interaction networks and the visual analysis of lateral gene transfer in bacterial phylogenetic trees. Here, we explain these projects and the main findings associated with the use of computational methods. Among the results are the evaluation of stability of ranking and signature methods applied to discovery proteomics data, a new approach to select candidate proteins from discovery to targeted proteomics, lists of candidate biomarkers for oral cancer, and new techniques for the visualization of biological networks and phylogenetic supertrees.