Visualisation and analysis of the dynamics and organisation of the yeast interactome

Download files
Access & Terms of Use
open access
Copyright: Goel, Apurv
Altmetric
Abstract
Protein-protein interaction networks are increasingly used for understanding the structure and dynamics of cellular function. They are an essential tool for systems biology. The data used to generate protein interaction networks is often combined from several sources. As this does not consider the conditions the interactions were measured in, it means that the resulting static networks do not correctly represent an organism at any particular time, place and/or condition. Dynamic networks can help alleviate this problem by incorporating further data, to take into account the changing nature of protein interactions, and thus the changing nature of interaction networks. In this Thesis, time-series protein abundance data (or gene expression as a proxy) was used to infer the existence of certain protein-protein interactions in networks. GEOMI, an open source Java-based network visualisation tool, was extended to map time-series abundance data onto networks. This formed network animations that changed dynamically, in accordance with experimental time. Using this capability, localised dynamic networks of all single interaction interface hub proteins in baker s yeast were generated and analysed. It was found that hubs could be classified by their dynamic binding behaviour, either as (i) competitive, if their partners compete for access to the single binding interface, or (ii) non-competitive if the expression of their partners is tightly controlled to minimise binding competition. Subsequently, this Thesis sought to use network-based analyses to better understand the basis of synthetic lethality in baker s yeast. Integrated intracellular networks were constructed, combining high confidence protein-protein interactions, kinase-substrate interactions and transcription factor target gene relationships. All triplet motifs in the networks, each of which contained one pair of synthetically lethal proteins, were then enumerated and analysed. This revealed that synthetic lethality does not arise at random; but was strongly over-represented (i) as part of multiprotein complexes (ii) when a kinase, its target, a transcription factor or its target had been duplicated in the cell (iii) where synthetic lethal pairs showed high positive correlation and periodic expression in the cell cycle and (iv) in pairs of proteins that showed low variability of abundance (and were thus under tight regulation in the cell).
Persistent link to this record
Link to Publisher Version
Link to Open Access Version
Additional Link
Author(s)
Goel, Apurv
Supervisor(s)
Wilkins, Marc
Bain, Michael
Creator(s)
Editor(s)
Translator(s)
Curator(s)
Designer(s)
Arranger(s)
Composer(s)
Recordist(s)
Conference Proceedings Editor(s)
Other Contributor(s)
Corporate/Industry Contributor(s)
Publication Year
2013
Resource Type
Thesis
Degree Type
PhD Doctorate
UNSW Faculty
Files
download whole.pdf 5 MB Adobe Portable Document Format
Related dataset(s)