Knowledge representation for data integration and exploration in translational ...

• Autor(es): Machado, Cátia Maria, 1981-
• Data: 2014
• Identificador Persistente: http://hdl.handle.net/10451/10846
• Origem: Repositório da Universidade de Lisboa
• Assunto(s): Representação do conhecimento; Ontologia; Interoperabilidade; Semântica de computadores; Biomedicina; Medicina translacional; Teses de doutoramento - 2014
• Tipo de Documento: Tese de Doutoramento
• Idioma: Inglês
• Orientador(es): Couto, Francisco José Moreira; Freitas, Ana Teresa

Descrição

Tese de doutoramento, Informática (Bioinformática), Universidade de Lisboa, Faculdade de Ciências, 2014 Biomedical research has evolved into a data-intensive science, where prodigious amounts of data can be collected from disparate resources at any time. However, the value of data can only be leveraged through its analysis, which ultimately results in the acquisition of knowledge. In domains such as translational medicine, data integration and interoperability are key requirements for an efficient data analysis. The semantic web and its technologies have been proposed as a solution for the problems of data integration and interoperability. One of the tools of the semantic web is the representation of domain knowledge with ontologies, which provide a formal description of that knowledge in a structured manner. The thesis underlying this work is that the representation of domain knowledge in ontologies can be exploited to improve the current knowledge about a disease, as well as improve the diagnosis and prognosis processes. The following two objectives were defined to validate this thesis: 1) to create a semantic model that represents and integrates the heterogeneous sources of data necessary for the characterization of a disease and of its prognosis process, exploiting semantic web technologies and existing ontologies; 2) to develop a methodology that exploits the knowledge represented in existing ontologies to improve the results of knowledge exploration methods obtained with translational medicine datasets. The first objective was accomplished and resulting in the following contributions: the methodology for the creation of a semantic model in the OWL language; a semantic model of the disease hypertrophic cardiomyopathy; and a review on the exploitation of semantic web resources in translation medicine systems. In the case of the second objective, also accomplished, the contributions are the adaptation of a standard enrichment analysis to use data from patients; and the application of the adapted enrichment analysis to improve the predictions made with a translational medicine dataset.