Lars Arvestad: A fast method for selecting phylogenetic replacement rate models
Time: Wed 2019-09-11 11.30
Location: Room 306, SU, Kräftriket
Lecturer: Lars Arvestad
Evolution is important in modern biology and medicine, and a phylogenetic tree can both give an enlightening explanation of molecular systems and serve as a foundation for further studies, experimental as well as computational. It is, unfortunately, still often difficult to infer a phylogenetic tree given sequence data. One important step in phylogenetic inference, which is too often overlooked, is to choose an appropriate transition matrix for a Markov process that describes the substitutions of the ”letters” in the input DNA or protein sequences.
It has become established practice to use maximum likelihood as model selection principle, which is computationally demanding: every plausible model (and sub-model) is tested, leading to an unfortunate combinatorial effect. Being able to quickly select an appropriate model, or at least reduce the set of models to test with maximum likelihood, would simplify experimentation.
I will in this talk describe the phylogenetic inference and model selection problems. A fast method for model selection that is based on the eigen decomposition of transition matrices will be proposed. I will also share some preliminary results on both simulated and biological data.