The network analysis has become a basic tool to gain insights on evolution and organization of living organisms in computational system biology. Since a group of genes may get involved into a biological process other than act alone, identifying modules from biological networks has been a central challenge to this field in the past decade. Several representative methods have been proposed to search such important modules using different intuitions while no unified framework exists yet, especially for multilayer networks, which can model gene expression dynamics and species conservation. This thesis provides a comprehensive study on active modules identification in multilayer intracellular networks, with the following main contributions: - An improvement on a heuristic method for identifying active modules from protein-protein interaction (PPI) networks. - A new objective of active modules to incorporate the topological structure and active property on the single layer and multilayer dynamic PPI network, and a convex optimization algorithm to solve it. - A new definition for active modules in single layer and multilayer gene co-expression networks and a novel algorithm which achieves the state-of-the-art performance. - A framework to conduct networks comparison via modules differentiation analysis, which can find condition-specific modules as well as conserved modules.

This thesis is presented in two volumes: the research (Volume One) and clinical (Volume Two) components. Volume One presents three research papers. The first paper is a systematic review exploring sensory modulation difficulties in rare genetic syndromes associated with Intellectual Disability and Autism Spectrum Disorder. The second is an empirical paper examining the behavioural phenotype in Pallister-Killian Syndrome (PKS). The third is a public dissemination report written to inform carers/parents of individuals with genetic syndromes with the results of the two papers. Volume two consists of five clinical practice reports (CPRs). CPR One presents two formulations (using behavioural and psychodynamic models) of a young girl with anxiety and a specific phobia of vomiting. CPR Two is a service evaluation determining if a neurodevelopmental pathway assessing ASD in young people met the standards of NICE guidelines. CPR Three is an experimental functional analysis of a young girl with Tuberculous Complex who was displaying aggressive behaviour. CPR Four is the case study, including assessment, formulation and intervention of a woman with schizoaffective disorder who required a relapse prevention plan. The abstract of CPR 5 (an oral case presentation) is the assessment of a man with Mild Cognitive Impairment, low mood and Leukaemia.

Many human cancers are more prevalent in men than women. This disparity is not fully explained by differences in key risk factor exposures, which suggests a possible genetic cause. Recent research has reported a link between loss of the Y chromosome (LoY) and increased incidence of non-haematological cancers. Using data from The Cancer Genome Atlas, I conducted an integrated, multi-‘omic analysis of Y chromosome methylation and copy number aberrations in three different cancers – colon, head and neck and kidney. My results indicate that aberrant methylation of the Y chromosome is common in all three cancer types. Hyper-methylation occurs in short, discrete regions, interspersed among wider regions of more general hypo-methylation. I also show that LoY is the most common aneuploidy in all three cancers, affecting between one third and one half of patients. Furthermore, both aberrant methylation and LoY are associated with reduced expression of potentially important genes. Most interestingly, for HPV negative head and neck cancer patients, I show a statistically significant association between LoY and worse survival, and that LoY may be linked to smoking. Subject to further validation, this suggests that LoY could be important in the pathogenesis of head and neck cancer for HPV negative patients.

Death Receptor 3 (DR3) is a death domain (DD) containing member of the Tumour Necrosis Factor Receptor Superfamily (TNFRSF) and has a single acknowledged TNFSF ligand called TNF-like protein 1A (TL1A). Previous research has implicated roles for DR3 in host immune defence and in various inflammatory diseases such as rheumatoid arthritis, inflammatory bowel disease (IBD), atherosclerosis and allergic lung inflammation. This thesis investigated a potential role for DR3 in co-ordinating the innate immune response, using an in vivo Staphylococcus epidermidis supernatant (SES) model of acute peritoneal inflammation. A further point of investigation looked at the effect of the absence of DR3 on thickening of the peritoneal membrane induced by repeated SES inflammation. My results showed that the DR3/TL1A pathway is not essential in maintaining the number of peritoneal or blood leukocytes during naive conditions. Stromal DR3 was found to be important in co-ordinating the innate immune response after the induction of acute SES induced inflammation, with significantly lower numbers of specific myeloid and lymphoid cell subsets accumulating in the peritoneal cavity of DR3 knockout (DR3-/-) mice. Despite this reduction in selected leukocyte numbers, the proportion of infiltrating cells exhibiting proliferation and cell death was unaffected by the absence of DR3. However reduced leukocyte numbers were associated with a significant reduction in the concentration of multiple chemoattractants in DR3-/- peritoneal supernatants compared to those from DR3+/+ mice. Quantitative RT-PCR data (qPCR) were consistent with the peritoneal membrane being a source of many of these chemoattractants. vi Results presented here identify for the first time a pro-inflammatory role for stromal DR3 in the innate immune response. However after repeatedly inducing inflammatory conditions DR3 promoted thickening of the peritoneal membrane, while an absence of DR3 prevented aberrant inflammation-induced tissue fibrosis.

Background: Controlled vocabularies in the molecular biology domain exist to facilitate data integration across database resources. One such tool is the Gene Ontology (GO), a classification designed to act as a universal index for gene products from any species. The Gene Ontology is used extensively in annotating gene products and analysing gene expression data, yet very little research exists from a library and information science perspective exploring the design principles, philosophy and social role of ontologies in biology. Aim: To explore how molecular biologists, in creating the Gene Ontology, devised guidelines and rules for determining which scientific concepts are included in the ontology, and the criteria for how these concepts are represented. Methods: A domain analysis approach was used to devise a mixed methodology to study the design of the Gene Ontology. Concept analysis of a GO term and a critical discourse analysis of GO developer mailing list texts were used to test whether ontological realism is a tenable basis for constructing objective ontologies. A comparison of the current GO vocabulary construction guidelines and a study of the reasons why GO terms are removed from the ontology further explored the justifications for the design of the Gene Ontology. Finally, a content analysis of published GO papers examined how authors use and cite GO data and terminology. Results: Gene Ontology terms can be presented according to different epistemologies for concepts, indicating that ontological realism is not the only way objective ontologies can be designed. Social roles and the exercise of power were found to play an important role in determining ontology content, and poor synonym control, a lack of clear warrant for deciding terminology and arbitrary decisions to delete and invent new terms undermine the objectivity and universal applicability of the Gene Ontology. Authors exhibited poor compliance with GO data citation policies, and in re-wording and misquoting GO terminology, risk exacerbating the semantic problems this controlled vocabulary was designed to solve. Conclusions: The failure of the Gene Ontology to define what is meant by a molecular function, the exercise of power by GO developers in clearing contentious concepts from the ontology, and the strict adherence to ontological realism, which marginalises social and subjective ways of classifying scientific concepts, limits the utility of the ontology as a tool to unify the molecular biology domain. These limitations to the Gene Ontology design could be overcome with the development of lighter, pluralistic, user-controlled ‘open ontologies’ for gene products that can work alongside more traditional, ‘top-down’ developed vocabularies.

Synucleinopathies are a group of diseases characterised by the presence of insoluble aggregated forms of α-synuclein. The most common of these diseases is Parkinson’s disease (PD) which affects approximately 1% of the UK population over the age of 60. Alpha-synuclein has also been linked to the disease through familial mutations and genome wide association studies as well as by its presence in sporadic cases. Although solid evidence exists for a role of α-synuclein in PD, it remains unclear as to how this protein exerts its toxicity on neurons and exactly how this leads to the cell death characteristic of this neurodegenerative disease. Alpha-synuclein belongs to a family of three proteins which also includes β- and γ-synuclein. These three proteins are highly homologous and evolutionarily conserved, however none of them have a well defined function. Evidence suggests a role for these proteins in synaptic vesicle dynamics but a more specific function remains to be unveiled. However, due to the considerable degree of homology across these three proteins, knockout models have been considered to allow functional compensation of the missing synuclein protein through one of the remaining family members. This has hindered studies from elucidating not only the role of α-synuclein but also β- and γ-synuclein. To overcome this problem triple synuclein knockout mice have been produced and characterised, as described in this thesis. As expected studies of these animals revealed no alterations in the number of dopaminergic neurons in either the substantia nigra pars compacta or ventral tegmental area. Despite this, a significant deficit in striatal dopamine concentrations was detected, regardless of the fact that the levels and function of tyrosine hydroxylase being normal. As well as this triple synuclein null mice were demonstrated to be hyperdopaminergic through various behavioural tests. Work employing physcostimulants and, through a collaboration, using fast scan cyclic voltametry suggested a role for these proteins in normal dopamine release dynamics at the level of the synaptic vesicle. A previous body of work has indicated that the loss of α- and/or γ-synuclein is able to provide a degree of resistance against the toxic affects of the dopaminergic neurotoxin MPTP. It was therefore hypothesised that the triple synuclein null animals would also display resistance to this toxin. However, these animals were shown to be more sensitive than wild type controls. Importantly it was apparent that animals lacking β-synuclein alone or in combination with other synucleins were the most sensitive to this toxin. Further work revealed a significant deficit in the ability of triple synuclein null mice to store dopamine in their synaptic vesicles. This may explain the sensitivity to MPP+, the active metabolite of MPTP, due to the fact it cannot be efficiently stored in synaptic vesicles, which restricts the toxins access to the mitochondria where it normally inhibits complex I, thus leading to cell death. When recombinant β-synuclein was reintroduced the deficit in synaptic vesicle dopamine uptake could be restored. However, β-synuclein can not do this alone and requires incubation with cytosolic factors, suggesting it acts as a chaperone in this role. This may explain why lines of synuclein null mice that specifically have the absence of β-synuclein apparently fair least well when exposed to MPTP. Finally, in order to assess the extent to which a loss of function role of α-synuclein leads to pathological alteration at the synapse an entirely novel conditional α-synuclein knockout mouse model was produced. Currently no ideal model exists to answer this question as conventional knockout models are based on the knockout of the protein in development. This may allow functional compensatory mechanisms to be established which can be overcome with a conditional knockout approach. As well as this it is important to assess this loss in an aged nervous system, as PD is a disease of aging. It is likely that, as α-synuclein forms insoluble Lewy bodies and undergoes abnormal posttranslational modifications, the amount of normally functioning protein at the synapse is depleted, therefore allowing a loss of function effect to develop. It is hoped this model will allow new insight into the early disease process. Overall this work further contributes to a body of evidence that suggests the synucleins play an important role in synaptic dopamine handling, particularly at the synaptic vesicle level. It is hoped that the newly established conditional α-synuclein knockout model will produce a new perspective on the loss of function role of α-synuclein in early disease development, an avenue that has yet to be fully explored.

ChIP-chip is a technology originally developed to determine the binding sites of proteins in chromatin on a genome wide scale. Its uses have since been expanded to analyse other genome features, such as epigenetic modifications and, in our laboratory, DNA damage. Datasets comprise many thousands of data points and therefore require bioinformatic tools for their analysis. Currently available tools are limited in their applications and lack the ability to normalise data so as to allow relative comparisons between different datasets. This has limited the analyses of multiple ChIP-chip datasets from different experimental conditions. The first part of the study presented here is bioinformatic, presenting a selection of tools written in R for ChIP-chip data analysis, including a novel normalisation procedure which allows datasets from different conditions to be analysed together, permitting comparisons of values between different experiments and opening up a new dimension of analysis of these datasets. A novel enrichment detection procedure is presented, suited to many formats of data, including protein binding (which forms peaks) and epigenetic modifications (which can form extended regions of enrichment). Graphical tools are also presented, to facilitate the analysis of these large datasets. A method of predicting the output of a ChIP-chip dataset is presented, which has been used to show that ChIP-chip is capable of detecting sequence dependent damage events. All functions work together, using a common data format, and are effcient and easy to use. The second part of this study applies these bioinformatic tools in a biological context. An analysis of Abf1 protein binding datasets has been undertaken, revealing many more binding sites than had previously been identified. Analysis of the sequences at these binding sites identifed the previously determined consensus binding motif in only a subset, with no novel motif identifiable in the remainder, suggesting binding may be in uenced by factors other than sequence.

This thesis examines the relationship between rights and duties in the field of fetal gene therapy and assesses if the current regulatory position within England and Wales is compatible with the intergenerational aspects of scientific progress within fetal gene therapy (FGT). Within the field of genomics, the fetal junction has become a site where gene therapists are developing a range of medical techniques, such as fetal gene therapy and in utero stem cell therapy. Utilising such techniques raises questions about the intergenerational aspects of scientific progress and how intergenerational rights can reshape regulation. The thesis focuses upon these key questions: Are the intergenerational issues of FGT taken into account by both direct and indirect stakeholders? Can intergenerational issues override the reproductive rights of the mother? Have intergenerational issues impacted upon the clinical applications implicit and manifest in this work? Addressing such questions is important because the conflict between the rights of the mother, fetus, clinical researchers and society have the potential to delay progress in FGT. In addressing these questions the thesis utilised thematic analysis of relevant regulatory institutional documents, from international declarations to regulatory guidelines; and semi structured interviews of identified FGT practitioners to identify areas of potential conflict. Following the data collection and analysis, the field data identified five key areas of potential conflict, which were then assessed using the Principle of Generic Consistency (PGC) as proposed by Alan Gewirth (1978) and later altered by Beyleveld and Brownsword (2001). The thesis will argue that the field data shows that established regulatory principles such as human dignity are of limited value in relation to FGT. In other areas such as informed choice, autonomy and intergenerational equity the PGC is applied to define and partially resolve the outstanding areas necessary for consistent ethical and regulatory guidance in FGT

The adenomatous polyposis coli (APC) gene encodes a multifunctional tumour suppressor protein that is essential for normal development. The most characterised role of APC is its ability to mediate the Wnt signaling pathway, a pathway disrupted in the majority of human cancers. The identification of a second adenomatous polyposis coli gene (APC2) which possesses many shared structural characteristics with APC and potentially comparable functions raises the possibility that APC2 also functions both in development and in tumour suppression, and that some redundancy may exist between the two proteins. Analysis of these proteins in the mouse has been hampered due to the lethality of the Apc mutation and the lack of a suitable Apc2 mutation. However, to circumvent the first of these difficulties, Cre-lox technology was employed to conditionally delete Apc in adult mouse tissues and so study its function in vivo. To circumvent the second difficulty, a novel Apc2 null allele had become available from the laboratory of Professor Hans Clevers. Remarkably, constitutive deletion of Apc2 does not lead to embryonic lethality, permitting study of the effects of Apc2 deficiency within adult tissues. In this thesis I aimed to characterise the consequences of Apc2 loss alone, and in the context of tissue specific Apc loss, in a range of tissues. Apc2 deficiency led to subtle changes in Wnt signaling in the intestines and liver however, no detectable differences of this pathway were apparent within the mammary gland. Phenotypically, altered homeostasis was only observed within the intestines. Apc2 deficiency led to an increase in epithelial cell division, an increase in markers of intestinal stemness and increases in intestinal cell migration. However, loss of Apc2 failed to induce tumourigenesis in the intestines or indeed any other tissue. In the context of Apc loss, the effect was dependent upon the tissue. Within the intestines, additional loss of Apc2 altered the immediate phenotype of Apc loss but failed to modify Apc induced tumourigenesis. Within the mammary gland, whilst either Apc protein alone was dispensable, combined loss synergised to disrupt homeostasis and drive tumourigenesis. Contrary to this, in the liver the additional loss of Apc2 attenuated tumourigenesis induced by reduced levels of Apc. Together, these studies highlight the importance of these proteins and their interactions and redundancies in homeostasis and tumourigenesis.

Many parasites of mammals avoid elimination by varying their exposed antigens. African trypanosomes—deadly parasites of humans and livestock in tropical Africa—possess a comprehensive system of antigenic variation (AV). Trypanosoma brucei undergo frequent, stochastic changes to their variant surface glycoprotein (VSG) coats, and therefore a developing immune response will be only partially effective against the trypanosome population as some trypanosomes will have already switched to a different VSG coat. The source of VSG variability is an archive of ~2000, mostly pseudogenic, silent VSG genes, of which only one is expressed. VSG genes can also be segmentally recombined: ‘mosaic’ VSGs, constructed from more than one silent VSG donor, allow both the reparation of pseudogenes and potentially generation of additional VSG variability. The aim of this research was to investigate the patterns of segmental VSG gene conversion in T. brucei, and assess its contribution to AV. Multiple, longitudinal samples were taken from chronic infections to follow the course of AV, and VSG cDNA sequences were analysed, building a detailed portrait of VSG expression across infection. VSG variability during an infection was extensive, and segmental gene conversion was found to be a frequent occurrence from approximately week three. Two main patterns were found: (i) expressed VSGs readily acquired a 3’ end different from their silent copy, a pattern that probably represents the 3’ boundary of gene conversion occurring within the coding sequence; (ii) expressed VSGs often appeared in sets of related ‘mosaics’, whereby more than one donor gene had contributed to the putative epitope-encoding part of the VSG. To test whether varying donor contributions represents an additional source of antigenic variability available to trypanosomes, a set of five mosaic VSGs retrieved from a single infection was expressed in non-switching trypanosomes and used to raise antibody responses. Indirect immuno-fluorescence, complement-mediated lysis, and agglutination assays using both polyclonal and monoclonal antibodies showed that although 4/5 mosaics were cross-reactive, one variant was completely antigenically distinct. Segmental gene conversion was therefore found to be both prominent in chronic African trypanosome antigenic variation, and capable of bringing antigenic novelty to an infection, with important consequences for the dynamics of AV, and the nature of selection pressure on the silent VSG archive.