Phylogenomic Analysis of Vertebrate Evolution

Abstract: The field of phylogenomics has emerged during the last five years, as the number of fully or partially sequenced genomes has increased rapidly. Using whole genomes for phylogenetic studies greatly increases the accuracy of the trees determined, and can potentially resolve phylogenetic questions that could not be resolved bu studies using only a few gene sequences. In the current thesis selected phylogentic problems in vertebrate evolution has been studies, with special emphasis on placental mammals. Sequences from several thousand protein-coding sequences (representing almost 3 million nucleotide characters) have been compared between over 30 species of placental mammals. The need for novel bioinformatical methods and approaches to work with such large amounts of data has been solved with customised software. The large-scale analyses has resolved several previously uncertain relationships with strong statistical confidence. However, a small number of divergences remained unresolved, despite the extensive data sampling. As other possibilities were ruled out as less likely, the natural explanation for this lack of resolution is probably to be found in speciation related processes. All the identified problematic relationships are characterised by rapid subsequent speciation events. In such scenarios, incomplete sorting of ancient polymorphism and species hybridization are processes that could create mosaic genomes in which different parts of the genomes will show incompatible trees. Divergences such as these are not well-represented as a classic bifurcating tree, instead attempts to illustrate them as phylogenetic networks have been made to more accurately describe the speciation.