Date of Award
Doctor of Philosophy (PhD)
Biochemistry and Molecular Biology
F. Alex Feltus
Leigh Anne Clark
Biological systems are incredibly difficult to untangle. On a molecular level, biological systems need to be analyzed by embracing high-dimensional genetic complexity in experimental design. One approach to understanding a biological system like a human organ (e.g., the normal brain) or disease state (e.g., brain tumor) is to identify condition-specific factors – biomarkers – that discriminate between biological states. A biomarker system is a group of biomarkers that are collectively associated with a phenotype. With exponentially growing amounts of high-throughput RNA-seq data available for both normal and different types of disease conditions, RNA-based biomarker systems, which underlie complex traits, can be identified. By using a combination of several computational biology approaches, including condition-specific gene co-expression networks (csGCNs) analysis, systems genetics integration of tissue-specific gene regulatory networks (tsGRNs), machine learning validation of biomarker systems, and dimensionality reduction techniques, robust RNA-based biomarker systems can be identified for both normal and disease states. In this Dissertation, I apply these methods to discover candidate biomarker systems involved in human normal brain region-specific states and normal lung versus lung tumor states. Chapter 1 provides overview of the field. Chapter 2 describes the identification of potential biomarker systems for normal human brain sub-regions by GCN network analysis. Chapter 3 shows the integration of csGCNs with lung-specific GRN to identify control-target biomarker systems for normal and cancerous lung tissue. Chapter 4 describes all other tissue-specific csGCNs constructed by KINC.
Hang, Yuqing, "Identification of Biomarker Systems in the Human Brain and Lung" (2021). All Dissertations. 2849.