Vanessa Jackson
MSc Biology

Maximum likelihood methods are commonly used to reconstruct ancestral character states on a phylogeny in order to test hypotheses about evolution and adaptation. My thesis project examines the accuracy of two maximum likelihood variants used to estimate discrete characters states on evolutionary trees. In the first method, termed "global", the transition-rate parameters for the likelihood model are estimated only once by maximizing over all states. In the second method, termed "local", a node is fixed in a certain state and then the transition rate parameters are estimated conditional upon that state. A custom software program was written to perform reconstructions on a variety of simulated phylogenies, as well as phylogenies of real biological organisms, in order to compare the accuracy of these methods. Preliminary results suggest that the global method performs better on marginal reconstructions, while the local method may perform better for joint reconstructions.

Courses taken outside of major
Mathematics - Computational Biology
Statistics - Computational Methods in
Statistics II (STA5107)

Florida State University