Date of Award

Fall 2022

Document Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Department

Computational Biology and Bioinformatics

First Advisor

Levine, Morgan

Abstract

Despite significant investment of resources for its study, Alzheimer’s Disease (AD) is an incurable disease with incredibly high prevalence in the older adult population. The confusion surrounding the recent approval of Aduhelm (Aducanumab) has highlighted a fraught history of inconsistent clinical trial results, difficult risk/ benefit analyses, and the desperate need for hope for an ever-increasing number of patients and their families. The primary motivation of this work is to directly characterize and embrace the heterogeneity of Alzheimer’s disease with the intent of future subtyping of patients in a data-driven manner. In this project we use extant postmortem multi-omics data from brain tissue to characterize the divergence between normal and diseased brain aging. To facilitate this, we describe a machine learning toolkit to advance the study of DNA methylation based aging predictors. First, we enable hypothesis driven approaches using a new R package for extant epigenetic clocks; Second, we describe the application of a new methodology, the PC Clocks framework, for the reduction of noise, particularly in brain tissue; Third, we present PCBrainAge, a brand new, robust predictor of brain aging that shows strong correlations with Alzheimer’s Disease. As this work uses postmortem brain tissue, we lay the groundwork for future capture of more homogeneous subtypes antemortem.

Download

Share

COinS