Examination Date

11-9-2023

Degree

Dissertation

Degree Program

Biostatistics

Examination Committee

Zhide Fanf, PhD; Hui_Yi Lin, PhD; Donald Mercante, PhD; Chih-yang Hu, ScD; Jian Li, PhD

Abstract

DNA methylation is one of the most well-known and well-studied epigenetic mechanisms that influences gene expression. Micro-RNA expression is also identified as playing a critical role in controlling gene expression. In recent years, with the development of high throughput technologies, the availability of multi-omics data has accumulated. As a result, obtaining data from micro-RNA (miRNA) expression, DNA methylation and gene expression on one sample become possible. Such availability offers us a great chance to run integrative analysis to get deep understanding of the regulating procedure.

Beta value is used for measuring the methylation level of the CpG sites. However, some researchers believe that instead of the position level, it is more reasonable to use the region level change to describe the methylation. Currently, this differentially methylated region (DMR) analysis is generally performed with linear models and certain clustering methods. While these methods work nicely and offered a lot of valuable discoveries, some potential alternative choices such as beta regression model methods might be available.

In this research, we propose to use beta regression model methods for DMR analysis and evaluate the performance of existing and our proposed DMR analysis methods. Besides, we will use the DMR analysis methods as well as mediation model to perform an integrative analysis to study possible relationships among miRNA-DNA methylation-gene expression profiles.

Download

Included in

Public Health Commons

Share

COinS