Análise da informação mútua em seqüências de DNA homólogas

• Ano de defesa: 2008
• Autor(a) principal: Pessoa, Luciana de Souza
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Programa de Pós-Graduação em Computação; Computação
• Palavras-chave em Português: Ciência da computação; Análise de Fourier; Teoria da informação; Biologia computacional; Análise de seqüências de DNA; Seqüências homólogas; CNPQ::CIENCIAS EXATAS E DA TERRA::CIENCIA DA COMPUTACAO::TEORIA DA COMPUTACAO::COMPUTABILIDADE E MODELOS DE COMPUTACAO
• Link de acesso: https://app.uff.br/riuff/handle/1/17873

In this dissertation, we describe a method for estimating the amount of mutual information of homologous DNA sequences, i. e., the information contained in a DNA sequence that can be used to identify homologous blocks that have a same ancestor. For that purpose, we using signal processing techniques, especially spectral analysis, signal filtering, and information theory. The analysis of the mutual information content allows us to estimate the probability of false positives strings that are not homologous to the given sequence, but are just as similar to it as the homologous ones. In 1999, Leitão and Stolfi developed an efficient algorithm for the reconstruction of fragmented seramic objects, using the technique of multi scale sequence comparison. This technique may be applicable to the problem of finding similar strings in a bio-sequence database, which is a fundamental problem for the identification of homologous genes and for the assembly of genomes from sequenced fragments. The viability of multi scale comparison for this problem relies on the hypothesis that, even in the coarsest scales used, a DNA sequence still contains enough mutual information to eliminate a significative fraction of false positives. We verify this hypothesis by the method describe here