Doctor of Philosophy in Bioinformatics

in the School of Biological Sciences

Shuting Lin

Defends her thesis:

Informatics Approaches for Identifying Drug-Specific Markers and Deciphering Genetic Regulation Mechanism in Cancer Treatment

Tuesday, Jan 28, 2025

11:30am Eastern

Krone Engineered Biosystems Building (EBB), Room #3029

Zoom link: https://gatech.zoom.us/j/92269946640

Thesis advisor:

Dr. Peng Qiu 
Department of Biomedical Engineering
Georgia Institute of Technology

Committee members:

Dr. Xiuwei Zhang, School of Computational Science and Engineering Georgia Institute of Technology

Dr. Soojin Yi, Department of Ecology, Evolution and Marine Biology University of California, Santa Barbara

Dr. Yong Teng, Department of Hematology and Medical Oncology Emory University

Dr. Saurabh Sinha, Department of Biomedical Engineering Georgia Institute of Technology 

Abstract: Advancements in cancer research rely on computational approaches to decipher genetic regulation and identify biomarkers that predict survival outcomes. This thesis focuses on developing informatics approaches to analyze large-scale genomic datasets, aiming to address key challenges in cancer bioinformatics and contribute to cancer treatment strategies. First, integrative analyses revealed biomarkers, including miRNAs and proteins, that are predictive of drug-specific survival outcomes. Many of these markers showed consistent relationships with their target or coding genes in terms of expression levels and correlations with survival outcomes. Validation through literature and independent datasets further underscored their significance. Next, the interplay between CpG methylation and alternative splicing was investigated. Correlation analyses between CpG sites, exons, and isoforms suggest that CpG methylation may influence alternative splicing by regulating exon inclusion or exclusion, consequently influencing isoform usage. These findings provide novel insights into the relationship between DNA methylation and transcriptional regulation. In addition, we developed prediction models to identify potential miRNA-gene interactions that had not been previously reported. External validation using independent datasets demonstrated that our approach is as robust as existing literature, highlighting its potential for uncovering new regulatory relationships. Finally, clustering analyses across different omics data identified distinct cancer patient clusters that were consistently associated with prominent survival outcomes, indicating robust and unique expression patterns across various genomic layers. Together, these findings provide a comprehensive framework for understanding cancer biology and supporting the development of precision oncology strategies to improve patient outcomes.