Author ORCID Identifier

0000-0002-1037-2698

Date of Award

Spring 5-8-2026

Document Type

Thesis (Ph.D.)

Department or Program

Cancer Biology

First Advisor

Brock Christensen

Abstract

Tumor epigenomes reflect a composite of malignant and non-malignant cellular signals, complicating interpretation of bulk DNA methylation data in heterogeneous tissues. In gynecologic cancers, where tumor progression, chromosomal instability, and microenvironmental variation are pronounced, failure to account for cellular composition obscures tumor-intrinsic regulatory remodeling and limits biomarker development. This dissertation advances a cell-type-aware epigenomic framework to disentangle specific DNA methylation alterations from compositional confounding and to extend methylation-based analysis to noninvasive biospecimens.

Using hierarchical reference-based deconvolution frameworks applied to large cohorts, cell-type-adjusted epigenome-wide analyses in endometrioid endometrial adenocarcinoma revealed coordinated epithelial-specific methylation remodeling associated with progression from early- to late-stage disease, enriched at regulatory elements and concordant with transcriptional reprogramming. In high-grade serous ovarian carcinoma, integration of structural genomic variation within diverse populations identified ancestry-associated differences and their potential contribution to tumor heterogeneity and disparities. Extending these methods beyond tumor tissue, DNA methylation-based cell typing of menstrual effluent demonstrated reproducible estimation of immune, epithelial, and stromal compartments across collection methods, supporting the feasibility of noninvasive epigenomic profiling for women’s health research.

Collectively, this work demonstrates that resolving cellular heterogeneity transforms bulk DNA methylation from a confounded composite signal into a scalable framework for interrogating tumor progression, structural genomic instability, and ancestry-associated variation. By integrating cell-type-resolved epigenomics across invasive and noninvasive contexts, this dissertation advances precision epigenomic approaches for gynecologic oncology and lays groundwork for equitable, noninvasive, population-scale biomarker development.

Original Citation

Vlasac, I. M., Stolrow, H. G., Thayer, Z. M., Christensen, B. C., & Rivera, L. (2025). DNA-based cell typing in menstrual effluent identifies cell type variation by sample collection method: toward noninvasive biomarker development for women’s health. Epigenetics20(1). https://doi.org/10.1080/15592294.2025.2453275

1_Supplementary_Table_genomic_enrichment_feature_segment_tests_vs_reference.csv (16 kB)
Genomic enrichment feature segment tests versus reference

2_Supplementary_Table_epithelial_genomic_enrichment_feature_segment_tests_vs_reference.csv (5 kB)
Epithelial-specific genomic enrichment feature segment tests versus reference

3_Supplementary_Table_promoter_both_genexp_DNAm_tbl.csv (6 kB)
Promoter-associated genes with concordant differential expression and DNA methylation

4_Supplementary_Table_enhancer_both_geneexpression_DNAm_tbl.csv (3 kB)
Enhancer-associated genes with concordant differential expression and DNA methylation

5_Supplementary_Table_GSEA_Advanced_vs_Early_cellAdj.csv (122 kB)
GSEA results: Late-stage versus Early-stage (cell-type adjusted model)

6_Supplementary_Table_epithelial_promoter_enhancer_comb_expression_and_DNAm.csv (161 kB)
Epithelial promoter and enhancer combined gene expression and DNA methylation results

7_Supplementary_Table_GSEA_Hallmark_Reactome_EpiInteraction_results.xlsx (40 kB)
GSEA Hallmark and Reactome epithelial interaction pathway results

Available for download on Wednesday, May 12, 2027

Share

COinS