Dissertation/Thesis Abstract

Identification of genetic factors in disease
by Below, Jennifer E., Ph.D., The University of Chicago, 2011, 234; 3460145
Abstract (Summary)

The field of genetic analysis has undergone a major shift in response to technological developments that have enabled massively high throughput genotyping and sequencing of genetic variants. The approaches used to identify genetic factors in disease rely on several factors: how well we understand and measure environmental cofactors to disease susceptibility, the complexity of the genetic model of the disease, and the quality and density of the genetic data. Each chapter of this thesis explores a different space within an analytic framework defined by these three measures of variation in complexity. First I explore two diseases, permanent neonatal diabetes and pancreatic agenesis which follow simple Mendelian inheritance patterns, are unaffected by environmental cofactors, and exhibit high quality genotype data on a sufficient number of SNPs leading to successful localization of the causative genetic factor. I then shift to a story of differential genotyping quality, in which for a subset of samples, the data generated to study the genetics of kidney disease in type 1 diabetes cases was of lesser quality than the rest. What follows is an example of genetic analysis in a disease with a known, quantifiable environmental susceptibility factor. I present two models of malignant mesothelioma. In the first, LINKAGE analysis followed by candidate gene resequencing identified the same causative gene in two families that are affected by a Mendelian form of mesothelioma, and in the second I present the work to-date in a population from Cappadocia, Turkey in which there is a mesothelioma epidemic. Finally, I present an example of my work from a more complex corner of this analytical framework. Type 2 diabetes is a disease known to have multiple (including heritable) environmental cofactors and a complex genetic model. By genotyping followed by imputation we assessed the genetic information at nearly 2 million genetic markers in an admixed population of Mexican Americans from Starr County, Texas, and followed the association analysis in this population with a meta analysis in a population from Mexico City, Mexico. To understand the functional roles of the top signals in these analyses, I annotated this data with the results of expression quantitative trait loci analyses in lymphoblastoid cell lines, muscle tissue, and adipose tissue. We identified a significant excess of SNPs that predict expression in muscle and adipose tissues among the top signals for type 2 diabetes in both Starr County and Mexico City. Collectively, these analyses demonstrate some of the breadth of current analytic problems faced by scientists attempting to identify genetic factors in disease. The methods, observations, and lessons learned from examples such as those presented here will be crucial to the success of genetic analysis as we move our focus toward identifying the effects of rare and even private mutations in exome and whole genome sequence data.

Indexing (document details)
Advisor: Cox, Nancy J.
Commitee: Abney, Mark, Ahsan, Habibul, Gilad, Yoav, Testa, Joseph
School: The University of Chicago
Department: Human Genetics
School Location: United States -- Illinois
Source: DAI-B 72/09, Dissertation Abstracts International
Subjects: Genetics, Pathology
Keywords: Diseases, Genetic variants, Genotyping, Mesothelioma, Pancreatic agenesis, Permanent neonatal diabetes
Publication Number: 3460145
ISBN: 9781124717289