Computational biology approaches for studying gene regulatory network discovery and modelling

Abstract

The advent of next generation sequencing has increased the gap between genome sequence data and knowledge, enhancing the need for faster means to fill this gap. The development of efficient computational biology methods to handle this gap has never been so important. Gene regulatory networks in particular have been studied widely for their role in controlling cellular behaviour, resulting in manifold phenotypic characteristics. In this thesis, I present novel techniques contributing to the discovery of gene regulatory network connections, through enhanced binding site prediction, binding site multiple sequence alignment and binding site specificity. Another major advantage of computational biology is the ability to simulate the behaviour of gene regulatory networks, in order to study the governing dynamics of such networks. In this thesis, I also introduce a new modelling language bringing computational modelling capabilities into the biological domain to simplify the process of writing a model that can be simulated in silico. I have proved through this work that: first, the devised computational biology techniques can provide cheap yet powerful and efficient techniques to study gene regulatory networks; and second, the techniques presented have novel superiority over current research in this domain.