UCLA

RNA expression profiling and DNA methylation analysis have been essential tools in understanding genomic mechanisms underlying human health and disease. Although many annotation databases are publically available, alternative data resources may be overlooked. This work focuses on the development of computational tools and strategies that incorporate results from both the leading functional annotation tools as well as working directly with publicly available expression and methylation datasets. Chapter 1 outlines the leading approaches for interpreting DNA methylation and RNA expression analyses. In addition, chapter 1 provides a brief background of Burkitt’s lymphoma and amyotophic lateral sclerosis (ALS) for studies discussed in later chapters. In chapter 2, we developed a set of methylation characterization and visualization tools for bisulfite sequencing data. These tools also characterize methylation levels at genomic features, like gene bodies as well as transcription factor targets. We provide a means to detect epigenetic regulation of transcription factor binding sites. Chapter 3 describes a multi- omics approach to understand an epigenetic mechanism for chemoresistance in Burkitt’s lymphoma. Burkitt’s lymphoma cell lines were cultured with drugs and developed increasing levels of resistance to chemotherapy. By analyzing transcriptional profiles of the chemoresistant cell lines with healthy B-cells at different stages of maturation as well as subsequent integration of DNA methylation and ChiP-Seq data from the chemoresistant cell lines, we were able to propose a novel mechanism of drug resistance in which E2a and PRC2 drive changes in the B- cell epigenome. In chapters 3 and chapter 4, we focused on the transcriptional and DNA methylation analysis of peripheral blood mononuclear cells (PBMCs) of patients affected with amyotrophic lateral sclerosis (ALS). Using transcriptional data of monocytes stimulated by different molecules, we were able to categorize our samples into inflammatory and non- inflammatory groups. A pathway enrichment analysis of the differentially expressed genes reveals potential targets of immune based treatments for ALS. In chapter 5, we investigated the differences in DNA methylation profiles in PBMCs from a pair of monzygotic twins discordant in the diagnosis for ALS. We developed a cell type abundance analysis method which suggest that the affected twin loses T-cells and gains monocytes during the course of the disease. Our direct use of reference data sets highlights the potential for understanding RNA-Seq and BS-Seq data and provides the groundwork for development of generalized transcription or methylation analysis tools, like CEllFi. Chapter 6 outlines the implementation of CEllFi, a bisulfite sequencing based method that allows for cellular deconvolution of heterogenous samples.