Statistical Methods for Integratively Characterizing Genetic and Genomic Data
Citation
Gaynor, Sheila Marie. 2018. Statistical Methods for Integratively Characterizing Genetic and Genomic Data. Doctoral dissertation, Harvard University, Graduate School of Arts & Sciences.Abstract
The era of multi-omics has provided the opportunity to explore complex genetic and genomic relationships. We consider approaches to analyzing such data integratively, towards understanding regulatory relationships and biological mechanisms. Chapter 2 seeks to elucidate the causal mechanisms of biological pathways leading to common, binary phenotypes. We introduce an estimator that allows for the identification of methylated sites mediating the association between smoking habits and airway obstruction in the Normative Aging Study. Chapter 3 considers network representations of expression quantitative trait loci (eQTL), and focuses on the network metric degree to assess how highly connected variants are within networks. We construct networks across tissues from the Genotype-Tissue Expression (GTEx) Project and make global comparisons of degree across eQTL networks to assess functionality and tissue-specificity of disease-related SNPs. Chapter 4 focuses on detecting communities of genes in eQTL networks, where spectral clustering can be used on the estimated network to identify structure. We characterize the impact of performing community detection on a regression-based network and assess the reproducibility of gene communities in GTEx eQTL networks using adjusted or standard spectral clustering.Terms of Use
This article is made available under the terms and conditions applicable to Other Posted Material, as set forth at http://nrs.harvard.edu/urn-3:HUL.InstRepos:dash.current.terms-of-use#LAACitable link to this page
http://nrs.harvard.edu/urn-3:HUL.InstRepos:41129144
Collections
- FAS Theses and Dissertations [5858]
Contact administrator regarding this item (to report mistakes or request changes)