Seeing the wood for trees: analysing multiple alternative phylogenies
Seminar Room 1, Newton Institute
Phylogenetic analysis very commonly produces several alternative trees for a given set of taxa. This can be due to having several, possibly conflicting, sources of information (such as different sets of orthologous genes) or as a result of the nature of the analytic method itself (for example, sampling a large number of trees from a posterior distribution). Faced with several alternative trees, biologists typically seek a single concensus phylogeny -- thereby discarding important information.
We review some recent approaches to dealing with alternative phylogenies that arise from a multivariate statistical view-point. We go on to describe a novel approach for comparing alternative phylogenies that is based on the idea of a "tree-of-trees": trees with similar topologies are clustered together in the same way that a single phylogeny clusters species with similar DNA sequences. The tree-of-trees is constructed in an analogous way to a most parsimonious tree for DNA data. We illustrate these ideas by applying our method to some simple data sets.
If it doesn't, something may have gone wrong with our embedded player.
We'll get it fixed as soon as possible.