Off-campus UNL users: To download campus access dissertations, please use the following link to log into our proxy server with your NU ID and password. When you are done browsing please remember to return to this page and log out.
Non-UNL users: Please talk to your librarian about requesting this dissertation through interlibrary loan.
Network-Based Multiomics Data Integration and Disease Marker Discovery Using Probabilistic Graph Modeling
Abstract
High-throughput technologies have revolutionized the ability to perform systems-level biology and elucidate molecular mechanisms of disease through comprehensive characterization of different layers of biological information. Integration of these heterogeneous layers can provide insight into the underlying biology but is challenged by modeling complex interactions. In this dissertation, we applied systems biology, network biology, and probabilistic graph modeling to integrate heterogeneous biological omics data and identify risk and disease signatures. We introduce two algorithms: 1) OBaNK: omics integration using Bayesian Networks and external knowledge, an algorithm to model interactions between heterogeneous high-dimensional biological data to elucidate complex functional clusters and emergent relationships associated with an observed phenotype. 2) DMA: data modification algorithm for multiomics data, an approach for system-level biomarker discovery that alters targeted signal values and relearns network structures to determine the molecules that drive the phenotype. OBaNK successfully improved the accuracy of learning interaction networks from data integrating external knowledge, identified heterogeneous functional networks from real data, and suggested potential novel interactions associated with the phenotype. DMA captured preestablished proxy biomarkers and key nodes impacted by the perturbation of these proxies. We applied DMA to multiomics pregnancy data, and the algorithm identified previously established signature markers of pregnancy. We generated and analyzed plasma and cerebrospinal fluid multiomics datasets for postoperative delirium. In addition to single omics analysis, we used OBaNK to identify mechanisms that could suggest delirium pathogenesis and pathophysiology. The proposed approaches model molecular interactions using probabilistic graph representations and provide complementary views to analyze multiomics data: how multifarious interactions explain underlying biological mechanisms and how the change in expression for key molecules describe the interactome for a phenotype. Future work would focus on extending OBaNK’s knowledgebase, testing alternative ways to select the groups of molecules for signal value alteration in DMA, and extending the delirium analysis to independent cohorts.
Subject Area
Bioinformatics|Biology|Immunology|Neurosciences
Recommended Citation
Tripp, Bridget Ann, "Network-Based Multiomics Data Integration and Disease Marker Discovery Using Probabilistic Graph Modeling" (2023). ETD collection for University of Nebraska-Lincoln. AAI30489205.
https://digitalcommons.unl.edu/dissertations/AAI30489205