Given a collection of organisms (or taxa), the objective of a phylogenetic analysis is to produce an evolutionary tree describing the genealogical relationships between the taxa. Inferring a phylogenetic tree that represents a good hypothesis of the truth is a difficult problem. Phylogenetic heuristics attempt to search a very small fraction of the exponentially-sized tree space for good approximations of the true evolutionary tree. The overall goal of the proposed research project is to develop information-rich methods for analyzing the large collection of trees encountered during a phylogenetic search. A better understanding of search behavior can drive the design of better heuristics, which ultimately lead to more accurate reconstructions of phylogenetic trees. Specific research objectives include: (i) developing a search history repository that contains all of the trees examined by a phylogenetic heuristic; (ii) designing algorithms for exploiting informative patterns in a search history repository; and (iii) developing visualization tools that provide informative views of the data residing in the repository. The novel data analysis and visualization techniques developed for this research project will complement existing NSF-funded efforts to build phylogenies such as ""The Tree of Life"", the evolutionary history of all known organisms. Moreover, the proposed work will provide mentoring and interdisciplinary training to a diverse group of students.
