Date of Award

Fall 1-1-2025

Document Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Department

Computational Biology and Bioinformatics

First Advisor

Higgins-Chen, Albert

Abstract

Biological aging unfolds along multiple heterogeneous trajectories shaped by both acquired molecular drift and inherited genetic variation. This dissertation investigates two complementary axes of aging heterogeneity: (1) stochastic, individually acquired epigenomic deviations; and (2) genotype-defined risk exemplified by the APOE?4 allele, the strongest genetic determinant of late-onset Alzheimer’s disease (AD). 1) Stochastic Epigenetic Mutation Detection We developed SEMdetectR, a robust R-based pipeline for identifying and quantifying stochastic epigenetic mutations (SEMs) -- rare, outlier CpG methylation events that accumulate idiosyncratically with age. Using technical replicates from three independent cohorts, we uncovered previous methodological flaws undermining SEM reliability and trained a machine learning-based filter that improves technical replicate concordance from <60% to >85%. Applied to the Framingham Heart Study, this enhanced SEM burden metric retains strong associations with mortality and cardiovascular disease, positioning SEMs as accessible biomarkers of personalized molecular drift. 2) APOE4?Driven Risk in Alzheimer’s Disease To capture genetically driven heterogeneity in brain aging, we analyzed multi-omic profiles from postmortem dorsolateral prefrontal cortex tissue in a cohort enriched for APOE4 carriers (n = 302 proteomes, n = 310 methylomes, n = 254 genomes). Among cognitively normal carriers, we identified early synaptic and metabolic perturbations. Among those with AD pathology, we uncovered APOE4-associated inflammatory and proteostatic upregulation and focal hypermethylation at ELAVL4. Mediation models show that a small protein module and a single CpG site mediate up to 30% of APOE4’s effect on AD pathology. Notably, proteomic and epigenetic biomarkers outperformed polygenic risk scores in stratifying AD status within APOE4 carriers. Together, these studies chart distinct yet complementary dimensions of aging heterogeneity -- stochastic epigenomic drift and genotype-dependent risk -- and deliver tools for their quantification. By enhancing the reliability of SEM detection and resolving molecular pathways specific to APOE4 carriers, this work advances the goals of precision geroscience: enabling early risk stratification and targeted interventions across diverse aging trajectories.

Download

Share

COinS