Title: New Methods for Reconstructing Very Large Evolutionary Trees
Monday 3/08 2pm-3pm
Speaker: Prof. Tandy Warnow
(University of Arizona, Faculty candidate interview talk)

Understanding the evolutionary history of a set of different species
or genes is an important objective for many biologists, and the study of
very large divergent datasets is becoming increasingly common. Many 
methods exist for reconstructing trees, and our understanding of
how well these methods perform under standard Markov models of
evolution has dramatically changed in the last few years.
I will describe these new results, and explain why they indicate
that the accurate recovery of very large evolutionary trees 
is unlikely to be accomplished through use of standard methods.
I will also describe a general "phylogenetic method booster", which 
improves accuracy of polynomial time methods, and decreases the 
computational effort in solving NP-hard optimization problems.