Informatics Approaches for Capturing Dynamic Gene Expression and Regulation

Author: Fangfang Yan, MS (2022)

Primary advisor: Zhongming Zhao, PhD

Committee members: Arif Harmanci, PhD; Lukas Simon, PhD; Hongyu Miao, PhD

PhD thesis, The University of Texas School of Biomedical Informatics at Houston.

ABSTRACT

Cellular systems are regulated through dynamic and complex mechanisms that involve various signaling and gene regulations. Aberrant changes may disrupt the system, leading to disease development. Time-series data analysis includes the analysis of datasets from multiple time points, which is essential to capture dynamic gene expression and regulation. Although many efforts have been put into the time-series gene expression analysis, limited studies have explored both gene expression and regulation in a time-series manner. Our study proposed the first time- specific informatics approach by integrating two critical regulators, TF and miRNA, and investigated their co-regulations. Specifically, we collected time-series bulk-tissue sequencing datasets covering mouse embryonic days (E) 10.5 to E14.5 and constructed novel regulatory networks between adjacent time points by assembling Feed-forward Loops (FFLs). We identified potential gene interactions that may influence orofacial development at different stages and conducted experimental validations of these networks in the related cell lines. Next, we applied the natural cubic spline model to better track the dynamic changes during the entire time course and identified eight clusters of genes with unique expression patterns. We then constructed pattern-specific regulatory networks by integrating data with CleftGeneDB, a database that manually curated orofacial cleft-related genes from literature. Lastly, to further resolve both gene expression and regulation dynamics at the cellular level, we simultaneously profiled single-cell transcriptomic and epigenomics from the mouse secondary palate across four developmental stages using novel single-cell multi-omic technology. Coupling differential gene expression analysis and motif enrichment analysis of positively linked peaks, we pinpointed a list of lineage-determining transcription factors, which were active at the early, middle, and late-stage of the trajectory, respectively. Collectively, our novel informatics approaches have advanced the understanding of molecular processes during orofacial development and the etiology of orofacial defects.