Finding the Missing Heritability:Gene Mapping Strategies for Complex Pedigrees.

Statistical Genetics;Genetics;Science;Human Genetics

Shah, Kaanan PradeepAntonellis, Anthony ;

University of Michigan

关键词: Statistical Genetics;    Genetics;    Science;    Human Genetics;   

Others  :  https://deepblue.lib.umich.edu/bitstream/handle/2027.42/99917/kaanan_1.pdf?sequence=1&isAllowed=y

瑞士|英语

来源: The Illinois Digital Environment for Access to Learning and Scholarship

【 摘 要 】

Geneticists have been working for decades to identify genetic factors that underlie variation in complex traits. Yet much of the variance attributed to additive genetic factors remains unaccounted for, the so-called ;;missing heritability problem.” Factors that may account for some of the missing heritability include the following: rare variants, structural variants, gene-gene interactions, and gene-environment interactions. In this dissertation, I evaluate the contribution of rare variants and gene-gene interactions to the missing heritability problem. Specifically, I develop and evaluate research strategies that take advantage of complex pedigree information. I apply these strategies to quantitative traits in the Old Order Amish, a population isolate in which most individuals are related through a single, complex pedigree. In Chapter 2, I describe a new statistical test to identify quantitative traits that are likely influenced by rare variants of large effect. I found evidence for the presence of rare variants influencing a few traits, including (remarkably) one for which a null mutation was previously identified. In Chapter 3, I evaluate the performance of Markov-chain Monte Carlo (MCMC) algorithms for linkage analysis of quantitative traits with complex pedigrees and dense genetic maps. I discovered that current algorithms fail to converge, resulting in highly variable LOD (logarithm of the odds) scores between MCMC runs. Despite this variability, I found consistent evidence of linkage for one trait for which a locus of large effect was previously mapped. Together, results from chapters 2 and 3 imply that rare variants of large effect are unlikely to explain much of the missing heritability of these traits. In Chapter 4, I consider that heritability might be overestimated rather than missing. To explore this possibility, I evaluate a new regression-based method to estimate heritability that is not inflated by gene-gene interactions. As suggested by Zuk et al. (2012), this method is ideal for use in population isolates but has not been investigated in realistic data settings. Unexpectedly, I discovered that the method produces biased estimates of the narrow-sense heritability, even for purely polygenic traits. Thus, caution should be exercised before using this method and attributing the missing heritability to gene-gene interactions.