My DMP Experience

Research

The goal of my research at TAMU this summer is to study the impact of starting trees on phylogenic search. In addition to known starting tree creation methods, I will be designing an algorithm for a novel starting tree construction. My hypothesis is that the resulting algorithm will fare better than starting trees generated by methods like neighbor-joining in maximum parisimony search.

An Overview of Phylogeny

Phylogeny is the study of how different organisms are related to each other. The impact of phylogeny can be found in various fields, including speices regeneration projects, drug research and discovery, as well as forensics. For example, a few years ago a Florida dentist was indicted on charges that he deliberately infected his ex-girlfriend with the HIV virus. They were able to prove his guilt by constructing a phylogeny of different strains of the virus. Phylogenies can be inferred in several ways; exhaustive search, while guaranteeing to find a tree that satisfies a certain criterion (for example, maximum parsimony), is often tedious and takes an unrealistic amount of time to complete. Tree space is very large; for 13 taxa, there are 13 billion different possible trees, or (2n-3)!!. For this reason, heuristic methods (which produce several good approximations of the "true" tree) have been used. These are known as phylogenetic searches.

What I'm trying to accomplish

In addition to the creation of the new algorithm, I will be comparing my method and known existing methods for starting tree creation against each other. By the end of the summer, I will have an open-source implementation of my code available for downloading, a technical report/research paper detailing the results of my research and my conclusions, as well as a scientific poster summarizing my results and methodology. In addition to the great learning experience, the other desired objective of this research project is to meet others at TAMU who are involved in phylogeny and its reconstruction, particularly from those whose research is more biological in nature.

© 2006 Suzanne J. Matthews | Design by Andreas Viklund