Meng Lin

Type

Text

Type

Thesis

Advisor

Eanes, Walter | Henn, Brenna M. | Kidd, Jeffrey | Rest, Joshua | Schatz, Michael

Date

2017-01-01

Keywords

Complex Trait | Evolution | Genetic Architecture | Human Genetics | Southern Africa

Language

en_US

Source

This work is sponsored by the Stony Brook University Graduate School in compliance with the requirements for completion of degree.

Identifier

http://hdl.handle.net/11401/78378

Publisher

The Graduate School, Stony Brook University: Stony Brook, NY.

Format

application/pdf | application/vnd.ms-excel

Abstract

The genetic architecture of complex traits likely differs across human populations. Yet the characterization of the underlying genetic variants in populations outside of Europe is far from adequate. This dissertation focuses on skin pigmentation as a highly heritable model phenotype, aiming to infer its complexity and adaptation in light skinned, endogamous, and genetically diverse KhoeSan groups from South Africa. Chapter 1 demonstrates the unexpectedly high polygenicity of pigmentation by assaying genome-wide array, exome, and high coverage resequencing data with paired quantitative pigmentation measurements. Both canonical and non-canonical pigmentation loci are associated with variation in the KhoeSan. Previously discovered pigmentation loci only explain a small fraction of the total phenotypic variance, but the largest effect variant (rs1426654) in SLC24A5 has a relatively high heritability of ∼15%. Chapter 2 details the evolutionary insight between the frequency of this SLC24A5 allele and the light skinned phenotype in KhoeSan. With simulations from a 4-population demographic model with selection, I show that the common haplotype is shared identically by descent with Europeans. Recent migration introduced the haplotype into southern Africa via ‘back to Africa’ gene flow about 2000 years ago. The frequency of the causal allele is consistent with strong subsequent positive selection, on the order of a ~0.05 selection coefficient. This example is among the strongest selective sweeps discovered in the human genome. In Chapter 3, I propose a novel approach to calculate statistical power when ancestry affects complex diseases. The simulation-based pipeline is developed as an easy and fast application for estimating statistical power in complex disease studies for admixed populations, assuming ancestry proportions and incidence rates. Together, the chapters provide a framework of human complex trait study set in a more complicated, admixed background. | 138 pages

Download

Share

COinS