A multimodal and multiobjective approach for phylogenetic trees reconstruction / Uma abordagem multi-objetivo e multimodal para reconstrução de arvores filogeneticas
AUTOR(ES)
Ana Estela Antunes da Silva
DATA DE PUBLICAÇÃO
2007
RESUMO
The reconstruction of phylogenetic trees can be interpreted as a systematic process of proposing an arborean description to the relative dissimilarities observed among sets of homologous genetic attributes of species being compared. The resulting phylogenetic tree presents a certain topology, or ancestrality pattern, and the length of the edges of the tree will indicate the number of evolutionary changes since the divergence from the common ancestor. Both topology and edge lengths are descriptive hypotheses of non-observable and conditional events, which implies the existence of diverse high-quality hypotheses for the reconstruction, as long as multiple performance criteria. This thesis (i) deals with unrooted trees; (ii) emphasizes the least squares, minimum evolution, and maximum likelihood criteria; (iii) proposes an extension to the Neighbor Joining algorithm which offers multiple high-quality reconstruction hypotheses; and (iv) describes and uses a new tool for multiobjective optimization in the context of phylogenetic reconstruction. Artificial and real datasets are considered in the presentation of results, which points to some advantages and distinctive aspects of the proposed methodologies
ASSUNTO(S)
arvores otimização matematica multiobjective otimization biologia - processamento de dados filogenia phylogenetic trees phylogeny reconstruction algorithms algoritmos geneticos minimum evolution
