Research
Figure 1: Phylogenetically corrected patterns of divergence of vertebral number in garter snakes (Thamnophis spp.) across three adaptive zones, estimated and compared with MIPoD.
Figure 2: Garter snake (Thamnophis elegans) in California.
Figure 3: Genome-wide Fst among replicate populations of threespine stickleback in Alaska. Vertical shading represents linkage groups (chromosomes), and solid arrows represent the most significant peaks of differentiation between ancestral oceanic and derived freshwater populations.
© 2007 Paul Hohenlohe
Research Program -- Overview
Since Darwin, evolutionary biology has faced a fundamental question: How do the microevolutionary processes operating within species scale up over long periods of time to explain macroevolutionary pattern? In particular, how does the genetic architecture of variation -- the raw material of evolution -- shape the structure of phenotypic variation, the response to selection, and the trajectories of diversification among taxa? I combine theoretical and empirical techniques to address this question:1) Phenotypic evolution
The theoretical foundation of evolutionary genetics provides a powerful framework for understanding microevolutionary process. In particular, it allows us to (1) analyze the genetic architecture of complex phenotypes in a statistical way, (2) empirically measure the form of the adaptive landscape, and (3) test predictive models of adaptation and divergence. Building on this foundation, I have developed a new class of phylogenetic comparative methods called MIPoD (Microevolutionary Inference from Patterns of Divergence). MIPoD leverages genetic information in the form of the G matrix in a comparative framework to rigorously test hypotheses about forms of natural selection and the adaptive landscape, based on actual patterns of divergence among related taxa. I am currently expanding the MIPoD approach to examine differences among adaptive zones (Figures 1 & 2) and to test alternative evolutionary process models of sexual selection (NSF co-PI with Steve Arnold) -- see some of our preliminary work here.
Related collaborative projects include simulation modeling to elucidate general patterns in the evolution of G, and a novel genetic network model to explore the relative roles of protein-coding versus cis-regulatory mutation, and to test the implicit assumptions of the G matrix in the context of genetic regulatory networks.
2) Molecular population genomics
Classical evolutionary genetics has taken a statistical and model-based approach to understanding genetic variation and evolutionary process, and its traditional connection to empirical data has been limited to a focus on one or a few genes or markers. However, with explosion of data from modern DNA sequencing technology, evolutionary biology is on the verge of uniting classical evolutionary genetics with a molecular and functional understanding of not just individual genes, but entire genomes.
With Bill Cresko and colleagues, I am developing novel analytical tools to estimate population genetic parameters in natural populations as continuous variables across the genome, and to conduct QTL and association mapping of phenotypic traits, using next-generation sequencing of RAD (Restriction site Associated DNA) markers. These techniques provide genomic sequence data at thousands of small regions across multiple individuals, without any prior development of markers, primers, or SNPs. From this work, the first high-density SNP-based genome scan of parallel adaptation in threespine stickleback will be out soon (Figure 3) -- we've found some remarkably consistent genomic patterns across independently derived populations, so stay tuned!
Because of the power of these techniques for association mapping and the emerging field of population genomics, I am collaborating with a number of colleagues on species of fish, snails, worms, and plants. In addition, I plan to apply them to the evolution of garter snake vertebral numbers (Figures 1 & 2), to connect the perspectives of quantitative genetics and population genomics.
Thanks to:
