Insights into heredity, heterogeneity, and homozygosity from new quantitative methods applied to large population genomic datasets

Nakka, Priyanka

Full Metadata

Overview

Year:: 2019
Contributor:: Nakka, Priyanka (creator); Rand, David (Reader); Weinreich, Daniel (Reader); Raphael, Benjamin (Reader); Ramachandran, Sohini (Advisor); Brown University. Biology and Medicine: Ecological and Evolutionary Biology (sponsor)
Genre:: theses
Subject:: Genetics; Computational biology
Extent:: 18, 166 p.
DOI: https://doi.org/10.26300/2rjn-8t88

Files

Description

Abstract:: Over the past decade, genome-wide associations studies (GWAS) and other methods have used human genetic data to successfully identify thousands of genomic elements underpinning complex phenotypes. However, the GWAS model makes several assumptions that need to be addressed to fully understand disease susceptibility in humans. For example, complex traits can be caused by multiple mutations of small effect and disease-causing variants do not always contribute additively to a phenotype or may also be subject to imprinting or other complex mechanisms. This dissertation presents multiple new approaches to aggregating genomic signals across genes and chromosomes to identify 1) loci underlying complex traits and diseases and 2) cases of uniparental disomy in large genomic datasets. First, I present a new method, called PEGASUS, for identifying genes underlying complex phenotypes using SNP-level GWAS p-values. Complex phenotypes may be caused by multiple mutations in a single gene or pathway, called “genetic heterogeneity”. By combining SNPs across genes while accounting for empirical linkage disequilibrium, we are able to more accurately identify associated genes in simulation and in studies for Attention-Deficit/Hyperactivity Disorder, Ulcerative Colitis and Waist-Hip Ratio. I also use PEGASUS to find novel genes and gene networks underlying ALL; in particular, we show that we are able to find networks containing genes that are known been shown to be implicated in leukemogenesis and hematopoiesis in mouse and zebrafish models, but to date have only been marginally significant in GWAS. Finally, I characterize the prevalence and phenotypic consequences of uniparental disomy (UPD), which is the inheritance of both homologs of a chromosome from one parent with no representative copy from the other. I use identity-by-descent (IBD) and our new machine-learning framework based on runs of homozygosity (ROH) to detect UPD and find that UPD occurs twice as commonly as previously thought (one in 2000 births) and a novel association between autism and UPD of chromosome 22. In short, this dissertation presents and applies multiple new methods to elucidate the complex relationship between human genetic variation and disease susceptibility in combination with large, genomic datasets.
Notes:: Thesis (Ph. D.)--Brown University, 2019

Content

Access Conditions

Rights: In Copyright
Restrictions on Use: Collection is open for research.

Citation

Nakka, Priyanka, "Insights into heredity, heterogeneity, and homozygosity from new quantitative methods applied to large population genomic datasets" (2019). Ecological and Evolutionary Biology Theses and Dissertations. Brown Digital Repository. Brown University Library. https://doi.org/10.26300/2rjn-8t88

Relations

Collection:

Ecological and Evolutionary Biology Theses and Dissertations

Theses and Dissertations for the Ecological and Evolutionary Biology department.
...