Bayesian Lasso for Detecting Rare Genetic Variants Associated with Common Diseases

Bayesian Lasso for Detecting Rare Genetic Variants Associated with Common Diseases

Bibliographic Details
Main Author: Zhou, Xiaofei
Language:English
Published: The Ohio State University / OhioLINK 2019
Subjects:
Statistics
Bayesian LASSO
common diseases
haplotype
rare variant
genetic analysis
combined design
survival analysis
GWAS
SNP
Online Access:http://rave.ohiolink.edu/etdc/view?acc_num=osu1563455460578675
id ndltd-OhioLink-oai-etd.ohiolink.edu-osu1563455460578675
record_format oai_dc
spelling ndltd-OhioLink-oai-etd.ohiolink.edu-osu15634554605786752021-08-03T07:11:54Z Bayesian Lasso for Detecting Rare Genetic Variants Associated with Common Diseases Zhou, Xiaofei Statistics Bayesian LASSO common diseases haplotype rare variant genetic analysis combined design survival analysis GWAS SNP Rare genetic variants are one of the key factors in understanding the etiology of common diseases. Although much focus on the literature has been on rare single nucleotide variants (rSNVs), rare haplotype variants (rHTVs) offer perhaps even greater biological relevance, as variants on the same chromosomes are often passed jointly. Several methods have been developed to detect rHTV effects on common diseases based on the Bayesian Lasso methodology for binary and quantitative traits (LBL) and greater power has been demonstrated over a number of rSNV-based methods. To further extend the capability of LBL, I work on two additional methods for detecting rHTVs associated with common diseases. The first method jointly analyzes independent case-control and family trio data, and is referred to as Combined Logistic Bayesian Lasso (cLBL). cLBL gains higher power than the single-design-based Bayesian Lassos because it achieves larger sample size by combining these two types of data. The likelihood used in cLBL is retrospective, making the method more efficient in identifying the rare haplotypes that are associated with diseases. The second method focuses on survival traits and is called Survival Bayesian Lasso (SBL). The current implementation of SBL is based on the accelerated failure time framework. SBL utilizes Weibull, loglogistic, and lognormal distributions to accommodate various types of potential hazards and interpretation schemes. A selection procedure is implemented in SBL to choose the most appropriate distribution. While SBL mainly focuses on rHTVs main effects, it can also evaluate environmental covariates as well as their interactions with HTVs. I applied SBL to The Cancer Genome Atlas breast cancer dataset and identified a risk rHTV that resides in the tumor suppressor CDH1 gene. I also conducted extensive simulations to gauge the performance of SBL. 2019-10-23 English text The Ohio State University / OhioLINK http://rave.ohiolink.edu/etdc/view?acc_num=osu1563455460578675 http://rave.ohiolink.edu/etdc/view?acc_num=osu1563455460578675 unrestricted This thesis or dissertation is protected by copyright: all rights reserved. It may not be copied or redistributed beyond the terms of applicable copyright laws.
collection NDLTD
language English
sources NDLTD
topic Statistics
Bayesian LASSO
common diseases
haplotype
rare variant
genetic analysis
combined design
survival analysis
GWAS
SNP
spellingShingle Statistics
Bayesian LASSO
common diseases
haplotype
rare variant
genetic analysis
combined design
survival analysis
GWAS
SNP
Zhou, Xiaofei
Bayesian Lasso for Detecting Rare Genetic Variants Associated with Common Diseases
author Zhou, Xiaofei
author_facet Zhou, Xiaofei
author_sort Zhou, Xiaofei
title Bayesian Lasso for Detecting Rare Genetic Variants Associated with Common Diseases
title_short Bayesian Lasso for Detecting Rare Genetic Variants Associated with Common Diseases
title_full Bayesian Lasso for Detecting Rare Genetic Variants Associated with Common Diseases
title_fullStr Bayesian Lasso for Detecting Rare Genetic Variants Associated with Common Diseases
title_full_unstemmed Bayesian Lasso for Detecting Rare Genetic Variants Associated with Common Diseases
title_sort bayesian lasso for detecting rare genetic variants associated with common diseases
publisher The Ohio State University / OhioLINK
publishDate 2019
url http://rave.ohiolink.edu/etdc/view?acc_num=osu1563455460578675
work_keys_str_mv AT zhouxiaofei bayesianlassofordetectingraregeneticvariantsassociatedwithcommondiseases
_version_ 1719456308692779008

Similar Items