Abstract:

 To construct a species tree, i.e. a tree describing the evolution of
 a given set of species, is a fundamental problem in
 computational biology. Often in practice, when a species tree is
 sought, a gene tree is constructed instead as a first good approximation.
 Since the two might differ significantly, a better approach is to
 infer the species tre not from one, but from a number of gene trees.
        In the talk we shall introduce the so-called duplication-loss
 model and see how it gives a solution to the above mentioned inference
 problem.

This is based on joint work with Mike T. Hallett.