Abstract:

The recent increase in availability of both sequence and expression data, along with ever-increasing computational power is making it possible to study evolutionary questions never before possible. I have applied both simulation and probabilistic modeling to questions of great interest in evolutionary genomics. In a collaborative project to study whole genome duplication in the Saccharomyces species complex, a major portion of our conclusions were dependent on a carefully developed simulation of the process of genome duplication and subsequent gene deletion. This simulation allowed us to conclude that this process was most likely neutral with respect to adjacent gene orientations. We were also able to conclude that genes were most likely deleted one at-a-time and not via long block deletions. It would not have been possible to draw either of these conclusions without our simulation. For another project to identify ectopie conversion events between duplicate genes from sequence alignment data, I developed a phylogenetic hidden Markov model. Not only does this probabilistic model outperform the most commonly used method, it is also capable of simultaneously locating and dating a conversion event within the alignment. In a final project, I describe another hidden Markov model that uses comparative tiling microarray data to identify regions of either copy number variation or expression variation between compared groups. I conclude this discussion with a short description of an application of this method to identify expression differences in two lines of Arabidopsis thaliana .