New Computational Methods to Study Historical Allopolyploidization in Plants

Committee Members

• Dr. Arun Sethuraman, Chair (SDSU)
• Dr. Michael Windham, Duke University
• Dr. Kate Ostevik, (UCR) 
• Dr. Jason Stajich (UCR) 
• Dr. Jeet Sukumaran (SDSU)

Abstract

This dissertation explores the evolutionary significance of whole genome duplication (WGD) in plant lineages. Chapter 1 presents the novel allotriploid Boechera lemmonil x paupercula x retrofracta, confirmed at eight sites in the central Sierra Nevada mountains of California.

Chapter 2 employs SpecKS, an original allopolyploid evolution simulator, to analyze the ancestral genetic diversity of over 200 plant lineages derived from ancient polyploidization. My experiments revealed that 95% of the lineages had high levels of ancestral diversity. SpecKS is the first Ks simulator to treat polyploidy as a two-dimensional continuum between auto- and allopolyploidy and genetic diversity.

In Chapter 3, I build on my prior work with SpecKS to create a more advanced polyploid evolution simulator called DemographiKS. I use DemographiKS to demonstrate the relationship between key demographic factors and the shape of the Ks distribution, revealing that the placement of the Ks peak is much more complex than previously understood, depending not on the time of whole genome duplication, but corresponds to the time of parental divergence (for allopolyploids) or the mean time of coalescence of the diploid ancestor (for autopolyploids under a population expansion). DemographiKS and SpecKS results are in agreement that the long tails observed on empirical polyploid Ks histograms are due to high levels of ancestral diversity.