A Bayesian Phylogenetic Mixture Model

Elisa Loza-Reyes, 
Dr. Anthony Robinson,
Dr. Merrilee Hurn

Department of Mathematical Sciences, 
University of Bath

Phylogenetic trees are the graphical structures which allow us to represent
the ancestral-descendant relationship between a group of evolutionary units
(e.g. populations, species, bacterial strains, genes) and to analyse it
statistically. Given a group of evolutionary units, we aim at approximating
the probability distribution functions of the parameters of interest, these
being the branching structure that relates them, the edge lengths of the
tree, the rates of nucleotide substitution.

We shall talk about the mathematical assumptions traditionally adopted in
likelihood-based methods of phylogenies reconstruction and go on to discuss
some of the proposals for relaxing these assumptions, which should lead to
more realistic ways of inferring phylogenies.