Identifiability of Phylogenetic Substitution Models
Elizabeth Allman (Mathematics and Statistics, University of Alaska)
(April 10, 2007 3:30 PM - 4:30 PM)
Molecular phylogenetics is concerned with inferring evolutionary relationships from biological sequences (for example, DNA or proteins). The probabilistic models of sequence evolution that underly statistical approaches in this field exhibit a rich mathematical structure.
After an introduction to the inference problem and an overview of phylogenetic substitution models, this talk will focus on the important statistical issue of identifiability of phylogenetic models. Identifiability is needed for valid inference of model parameters --- whether tree topologies, edge lengths, or substitution rates --- but current understanding still lags well behind common practice.