Use this URL to cite or link to this record in EThOS:
Title: Network theoretic tools in the analysis of complex diseases
Author: Banerji, C.
ISNI:       0000 0004 7428 8879
Awarding Body: UCL (University College London)
Current Institution: University College London (University of London)
Date of Award: 2015
Availability of Full Text:
Access from EThOS:
Full text unavailable from EThOS. Please try the link below.
Access from Institution:
In this thesis we consider the application of network theoretic tools in the analysis of genome wide gene-expression data describing complex diseases, displaying defects in differentiation. After considering the literature, we motivate the construction of entropy based network rewiring methodologies, postulating that such an approach may provide a systems level correlate of the differentiation potential of a cellular sample, and may prove informative in the analysis of pathology. We construct, analytically investigate and validate three such network theoretic tools: Network Transfer Entropy, Signalling Entropy and Interactome Sparsification and Rewiring (InSpiRe). By considering over 1000 genome wide gene expression samples corresponding to healthy cells at different levels of differentiation, we demonstrate that signalling entropy is a strong correlate of cell potency confirming our initial postulate. The remainder of the thesis applies our network theoretic tools to two ends of the developmental pathology spectrum. Firstly we consider cancer, in which the power of cell differentiation is hijacked, to develop a malicious new tissue. Secondly, we consider muscular dystrophy, in which cell differentiation is inhibited, resulting in the poor development of muscle tissue. In the case of cancer we demonstrate that signalling entropy is a measure of tumour anaplasia and intra-tumour heterogeneity, which displays distinct values in different cancer subtypes. Moreover, we find signalling entropy to be a powerful prognostic indicator in epithelial cancer, outperforming conventional gene expression based assays. In the case of muscular dystrophy we focus on the most prevalent: facioscapulohumeral muscular dystrophy (FSHD). We demonstrate that muscle differentiation is perturbed in FSHD and that signalling entropy is elevated in myoblasts over-expressing the primary FSHD candidate gene DUX4. We subsequently utilise InSpiRe, performing a meta-analysis of FSHD muscle biopsy gene-expression data, uncovering a network of DUX4 driven rewired interactions in the pathology, and a novel therapeutic target which we validate experimentally.
Supervisor: Severini, S. ; Teschendorff, A. ; Zammit, P. Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available