Computational methods for identifying cell type, signallying pathways and gene regulatory networks

Abstract

Due to the advancement of genetic technologies it is now possible to accurately and simultaneously measure the expression levels of essentially all genes for species that have had their genomes sequenced. A detailed molecular understanding of these gene expression data will facilitate identification of new drug and cell therapy targets for disease treatment. There are increasing amounts of genome-wide public gene expression data that provide valuable information for the research community. The underlying hypothesis for this thesis is that development of new bioinformatics methods for both cell type and signal pathway characterization - that make use of public gene expression data - will facilitate application of stem cells and their differentiated derivatives. Each chapter in this thesis had specific aims. Chapter 1 presents a literature review that aimed to overview current bioinformatics tools applicable to robust definition of stem cell and differentiated cell identity. Chapter 2 aimed to develop a new tool (termed C3) for identifying unknown cell types based on their transcriptional profile. Chapter 3 aimed to characterize a new populations of purified cells (termed ROR1+ cells) obtained from differentiation of human pluripotent stem cells. Chapter 4 aimed to develop a new method (termed SPAGI) for predicting active signalling pathways for any cell type, based on the cell's transcriptome. Chapter 5 aimed to provide a large-scale prediction (or blueprint) of signal pathway-mediated regulation of lens epithelial cell gene expression. Lastly, the final chapter provides an overall reflection on the thesis combined with perspectives on future work made possible as a result of the thesis.

Date of Award	2018
Original language	English