Abstracts (first author)
Whole-genome methylation patterns in the Khomani Bushmen of South Africa
The Khoe San hunter-gatherers of southern Africa are believed to be one of the world's most ancient populations based on their high genetic diversity and low levels of LD. Phenotypically, the Khomani Bushmen have a range of lighter skin pigmentations, intermediate between European and central African populations, as well as a typically shorter stature. We have collected genetic and phenotypic information from the Khomani Bushmen. Moreover, we methyl-typed these individuals at over 400,000 CpGs sites as well as genotyped them at over 300,000 SNP sites. In order to better understand the epigenetic contribution of DNA methylation to skin color, we use this data set to study genome-wide methylation profiles for skin pigmentation. We also study the relationship of DNA methylation and age in these individuals. These results can help us better understand the role of DNA methylation in the aging process and in shaping the human phenotype.
Evolution with stochastic epigenetic variation: a role for recombination
Phenotypic adaptation to fluctuating environments has been an important focus in the population genetic literature. Previous studies have shown that evolution under temporal variation is determined not only by expected fitness in a given generation, but also by the degree of variation in fitness over generations; in an uncertain environment, alleles that increase the geometric mean fitness can invade a randomly mating population at equilibrium. This geometric mean principle governs the evolutionary interplay of genes controlling mean phenotype and genes controlling phenotypic variation, such as genetic regulators of the epigenetic machinery. Thus, it establishes an important role for stochastic epigenetic variation in adaptation to fluctuating environments: by modifying the geometric mean fitness, variance-modifying genes can change the course of evolution and determine the long-term trajectory of the evolving system. The role of phenotypic variance has previously been studied in systems in which the only driving force is natural selection, and there is no recombination between mean- and variance-modifying genes. We have developed a population genetic model to investigate the effect of recombination between mean- and variance-modifiers of phenotype on the geometric mean principle under different environmental regimes and fitness landscapes. We show that interactions of recombination with stochastic epigenetic variation and environmental fluctuations can give rise to complex evolutionary dynamics that differ from those in systems with no recombination.