Using DNA sequences to infer the characteristics of populations

3 March 2006

John Wakeley
OEB, Harvard


Historical processes and events can be inferred from the patterns of genetic variation in a sample of DNA sequences or other genetic data. The gene genealogy of a sample is the pattern of ancestral genetic relationships among the member of the sample. It has proved very useful to invoke gene genealogies in trying to understand the forces that produce and maintain genetic variation and in constructing a framework for making inferences from genetic data. Ancestral processes are used in population genetics to generate priors for gene genealogies at a locus. Kingman's coalescent is by far the most commonly used ancestral process, yet it depends on a number of assumptions which may not be true for a given species. I will focus on an assumption about the distribution of the number of offspring per individual in the population. I will review Kingman's coalescent, then consider alternative models in which the variance of offspring number can be much larger than in Kingman's coalescent. I will present results for one particular such model, which can produce ancestral processes that differ dramatically from Kingman's coalescent, and discuss the application of the model to some data from the Pacific oyster.

current theory lunch schedule