ingest cdrApp 2018-08-23T17:02:08.718Z d39a25df-af15-48e9-aec2-c9af81a997a2 modifyDatastreamByValue RELS-EXT fedoraAdmin 2018-08-23T17:02:59.396Z Setting exclusive relation addDatastream MD_TECHNICAL fedoraAdmin 2018-08-23T17:03:10.854Z Adding technical metadata derived by FITS addDatastream MD_FULL_TEXT fedoraAdmin 2018-08-23T17:03:24.996Z Adding full text metadata extracted by Apache Tika modifyDatastreamByValue RELS-EXT fedoraAdmin 2018-08-23T17:03:47.094Z Setting exclusive relation modifyDatastreamByValue MD_DESCRIPTIVE cdrApp 2018-09-27T19:25:59.188Z modifyDatastreamByValue MD_DESCRIPTIVE cdrApp 2019-03-21T20:23:08.060Z Robert Corty Author Curriculum in Bioinformatics and Computational Biology School of Medicine VARIANCE HETEROGENEITY IN GENETIC MAPPING Genetic mapping is a process by which researchers seek to identify genetic factors that influence a trait of interest. Such efforts typically focus on those that either increase or decrease the trait of interest, and assume that the variance of the trait is constant across all individuals. I develop and apply statistical methods that challenge that assumption in two ways. First, I consider the situation where non-genetic factors influence trait variance, which I term “background variance heterogeneity”. Though they are not of immediate interest in a genetic mapping study, they can be exploited to align observations’ weights with their precisions. Second, I consider the situation where genetic factors influence trait variance, which I term “foreground variance heterogeneity”. Such factors are of immediate interest because they represent novel discoveries that could be missed by standard analyses. I consider both foreground and background variance heterogeneity as they relate to linkage disequilibrium mapping in exchangeable mapping populations. I report three novel genetic factors with strong evidence that they influence medically-important traits in the mouse model system. Finally, I consider the background variance heterogeneity as it relates to association mapping in non-exchangeable populations. I report a mathematical advance that makes possible the fitting of a statistical model that accommodates background variance heterogeneity in non-exchangeable populations. Summer 2018 2018 Genetics genetic mapping, qtl, variance heterogeneity, vqtl eng Doctor of Philosophy Dissertation University of North Carolina at Chapel Hill Graduate School Degree granting institution Bioinformatics and Computational Biology William Valdar Thesis advisor Fernando Pardo Manuel de Villena Thesis advisor Lisa Tarantino Thesis advisor Yun Li Thesis advisor James Evans Thesis advisor text Robert Corty Creator Curriculum in Bioinformatics and Computational Biology School of Medicine VARIANCE HETEROGENEITY IN GENETIC MAPPING Genetic mapping is a process by which researchers seek to identify genetic factors that influence a trait of interest. Such efforts typically focus on those that either increase or decrease the trait of interest, and assume that the variance of the trait is constant across all individuals. I develop and apply statistical methods that challenge that assumption in two ways. First, I consider the situation where non-genetic factors influence trait variance, which I term “background variance heterogeneity”. Though they are not of immediate interest in a genetic mapping study, they can be exploited to align observations’ weights with their precisions. Second, I consider the situation where genetic factors influence trait variance, which I term “foreground variance heterogeneity”. Such factors are of immediate interest because they represent novel discoveries that could be missed by standard analyses. I consider both foreground and background variance heterogeneity as they relate to linkage disequilibrium mapping in exchangeable mapping populations. I report three novel genetic factors with strong evidence that they influence medically-important traits in the mouse model system. Finally, I consider the background variance heterogeneity as it relates to association mapping in non-exchangeable populations. I report a mathematical advance that makes possible the fitting of a statistical model that accommodates background variance heterogeneity in non-exchangeable populations. Genetics genetic mapping; qtl; variance heterogeneity; vqtl Doctor of Philosophy Dissertation University of North Carolina at Chapel Hill Graduate School Degree granting institution Bioinformatics and Computational Biology William Valdar Thesis advisor Fernando Pardo Manuel de Villena Thesis advisor Lisa Tarantino Thesis advisor Yun Li Thesis advisor James Evans Thesis advisor 2018 2018-08 eng text Robert Corty Creator Curriculum in Bioinformatics and Computational Biology School of Medicine Genetic mapping is a process by which researchers seek to identify genetic factors that influence a trait of interest. Such efforts typically focus on those that either increase or decrease the trait of interest, and assume that the variance of the trait is constant across all individuals. I develop and apply statistical methods that challenge that assumption in two ways. First, I consider the situation where non-genetic factors influence trait variance, which I term “background variance heterogeneity”. Though they are not of immediate interest in a genetic mapping study, they can be exploited to align observations’ weights with their precisions. Second, I consider the situation where genetic factors influence trait variance, which I term “foreground variance heterogeneity”. Such factors are of immediate interest because they represent novel discoveries that could be missed by standard analyses. I consider both foreground and background variance heterogeneity as they relate to linkage disequilibrium mapping in exchangeable mapping populations. I report three novel genetic factors with strong evidence that they influence medically-important traits in the mouse model system. Finally, I consider the background variance heterogeneity as it relates to association mapping in non-exchangeable populations. I report a mathematical advance that makes possible the fitting of a statistical model that accommodates background variance heterogeneity in non-exchangeable populations. Genetics genetic mapping; qtl; variance heterogeneity; vqtl Doctor of Philosophy Dissertation University of North Carolina at Chapel Hill Graduate School Degree granting institution William Valdar Thesis advisor Fernando Pardo Manuel de Villena Thesis advisor Lisa Tarantino Thesis advisor Yun Li Thesis advisor James Evans Thesis advisor 2018 2018-08 eng text Corty_unc_0153D_17987.pdf uuid:9ed09d4a-57aa-4b47-9093-a8e700f1006f 2020-08-23T00:00:00 2018-07-07T02:38:12Z proquest application/pdf 11191862