Uma nova abordagem para a identificação de ilhas genômicas em bactérias com base no método de agrupamento mean shift

• Ano de defesa: 2017
• Autor(a) principal: Brito, Daniel Miranda de
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal da Paraíba; Brasil; Informática; Programa de Pós-Graduação em Informática; UFPB
• Palavras-chave em Português: Ilhas genômicas; Predição; Análise genômica; Agrupamento; Mean shift; CIENCIAS EXATAS E DA TERRA::CIENCIA DA COMPUTACAO
• Link de acesso: https://repositorio.ufpb.br/jspui/handle/tede/9040

Genomic islands (GIs) are regions of the bacterial and archaeal genomes that were acquired through the phenomenon of horizontal transfer. Usually, these regions provide important adaptations to these organisms, such as antibiotic resistance and pathogenicity, whose effects can be harmful to other species. For these reasons, many computational methodologies have been proposed for their prediction, however, none of them are capable to precisely identify the whole repertoire of islands present in a given genomic sequence. Therefore, the development of new approaches that explore different aspects of these regions is timely, allowing the identification of those not known. In this paper, it is proposed a novel method for the identification of GIs, built based on mean shift clustering algorithm, with the automatic bandwidth calculation, necessary to its operation. Test results with genomic island inserted in bacterial genomes show that the method is capable of identifying these regions, with sensitivity rates above 99%. Tests performed with bacterial genomes with known GIs revealed the potential of the method for their identification and for the discovery of new island. The detailed study of the new islands content pointed the presence of typical GIs elements, confirming its effectiveness in the prediction of these regions.