Search

The corrosion behaviour of thermally-sprayed sacrificial coatings was investigated under different operating conditions which simulated those of an offshore structure in active service. Electrochemical measurements, combined with several characterisation techniques, such as SEM, EDX and Raman spectroscopy were used to assess the performance of the coatings during exposures to various conditions of simulated full and alternating seawater immersion, splash zone conditions and atmospheric corrosion conditions. The effect of temperature and graphene-containing sealants were also studied under artificial seawater immersion on aluminium-coated steel samples containing defects. The results of immersion tests of thermally sprayed aluminium (TSA) coatings with defects revealed significantly worse performance of the coatings exposed to cold temperatures (4C) compared with other temperatures tested. A significant amount of rust was observed in the defect region and required polarisation (-0.8 V vs Ag/AgCl) was not provided by the TSA. This was related to low solubility of aluminium and decreased precipitation of calcareous deposits. The application of graphene - containing sealants on TSA coatings turned out to be effective in decreasing the contact between the electrolyte and the steel substrate at warm temperatures (25C) under full immersion conditions. However, the addition of the sealants was found to be detrimental to the performance of TSA coatings operating at cold temperatures (4C) under full immersion conditions as well as warm temperatures (25C) under alternating immersion conditions. The results of atmospheric corrosion tests showed that cathodic and anodic regions were not easily distinguishable on TSA coatings. Droplets did not spread, and sulphur-containing corrosion products were detected after the test. This highlights the importance of using seawater, and not NaCl solutions, as a corrosive medium simulating marine environment.

Plantation woodland is increasing in extent globally and as a proportion of global forest cover. Plantation woodland is not only an important element of the U.K. economy, but also is an important woodland habitat in the U.K., given the pre-historic and historic loss of woodland cover. Woodland policy aspirations in the U.K. are currently heavily focused on increasing woodland extent, coupled with maximising the ecosystem services delivered by these habitats, including the provision of renewable energy developments and increasing the provision of biodiversity services. Such broad management objectives often necessitate trade-offs, where management to deliver different desired outcomes conflict. This thesis focuses on the conservation management of European nightjar Caprimulgus europaeus and is intended to inform the development of evidence-based conservation interventions that can support nightjar and the moth species on which they feed, in plantation coniferous woodland and in particular in the Welsh Government Woodland Estate. Specifically, I use European nightjar life history parameters (e.g. nest success) and movement data, to explore nightjar ecology in areas of changing land use and management. In order to inform subsequent nest success analysis a comparison of the observed nesting success of tagged birds and untagged birds, whilst controlling for the potential confounding effects of weather, is used to explore and confirm the lack of observed tag effects on measures of nest success (e.g. fledging success) in nightjar breeding at Brechfa Forest Wind Farm (South Wales, UK). Moth diversity and biomass data from open habitats within forests across a gradient of ages, is then used to confirm the importance of native vegetation primary productivity and stand management in driving moth biomass and diversity. Moth biomass is then combined with GPS tag movement data to explore nightjar foraging behaviour within the forest matrix. This analysis confirms that nightjar movement is driven by spatial and temporal variation in the availability of moth biomass. Nest record data from sites across Wales is then used to explore the potential for wind farm construction disturbance effects on nest success (e.g. fledging success). Alongside this, I also examine the role of habitat management and foraging habitat availability. Overall, my results show that forest management can have significant effects on both nightjar nest success, movement and food (moth) availability. I also show that forest management has an important effect on the presence or absence of moths of conservation concern. These results also confirm that modifications to the spatial and temporal patterns of forest management can be used to support the conservation of both moths and nightjar, with likely cascading benefits for other aerial insectivores.

Under the circumstance of urgent demand for clean, renewable hydrogen fuel on a global scale, large-scale pure hydrogen production has drawn extensive attention from both industrial and science communities. To date, most commercial hydrogen is produced by steam methane reforming, which operated under harsh conditions with high energy-cost.1 Alternatively, electrolysis of water is an effective strategy that can produce high purity H2 under mild conditions. In this context, high active, low-cost electrocatalysts tend to be the specific target to realize the industrial application. To design and develop prospective catalysts, insight into the correlation between composition, coordination environment and electronic structure of electrocatalysts is undoubtedly important. With this in mind, this thesis reports various kinds of electrocatalysts that can effectively catalyze the hydrogen/oxygen evolution reaction (HER/OER), with tailored coordination structure/electronic structure. Through a variety of characterizations, the correlation between chemical environments of actives species and corresponding electrochemical activity towards HER/OER is elucidated. Chapter 4 presents the phase transformations from lepidocrocite titania nanosheets (L-TiO2) to rutile (R-TiO2) and anatase (A-TiO2).2 We report that temperature determined the final phase-structure in the transition phase of the L-TiO2 nanosheets into TiO2 nanoparticles, while the pH determined the final morphology and particle size. Based on the experimental results, two different transition pathways of dissolution-recrystallization (pH=3) and topologically rolling transition (pH=9) have been proposed. Atomic dispersed Pt species were deposited on the pre-synthesized A-TiO2 surface via photochemical reduction reaction. Due to the strong interaction of deposited Pt species and TiO2 supports, a lower oxidation state of Pt was found on Pt@TiO2-pH9. In 0.5 M H2SO4, Pt@TiO2-pH9 delivered the best HER activity than other Pt decorated TiO2 supports. It only requires an overpotential of 30 mV to reach 10 mA cm-2 geo. In addition, with regard to the mass activity, all the Pt decorated TiO2 samples show excellent HER performance superior to commercial Pt/C due to ultralow loading of Pt on the catalyst. Chapter 5 describes that low-loading (< 20 wt.%) Ni-confined in layered metal oxides anode catalysts have been synthesized by facile ion exchange methodology and subjected to systematic electrochemical studies.3 It was found that Ni-intercalated on K-rich birnessite (Ni-KMO) presents an onset overpotential (ηonset) as low as 100 mV and overpotential at 10 mA cm-2 (η10) of 206 mV in pH=14 electrolyte. By combining electrochemical methods XAS and XES, we demonstrate that Ni sites are the active sites for OER catalysis and that the Mn3+ sites facilitate Ni intercalation during the ion-exchange process. Different coordination structures of Ni species were confirmed on Ni-KMO and Ni-HMO, which are strongly correlated to the electrochemical behavior and corresponding OER activity. The effect of the pH and the nature of the supporting electrolyte on the electrochemical performance were also evaluated to reveal the electrochemical features of Ni intercalated KMO. Chapter 6 shows the surface galvanic formation of Co-OH on KMO was achieved via the redox reaction of hydrated Co2+ with crystalline Mn4+.4 The synthesis method takes place at ambient temperature without using any surfactant agent or organic solvent, providing a clean, green route for the design of highly efficient catalysts. The redox reaction resulted in the formation of ultrathin Co-OH nanoflakes. XAS and XPS analysis confirmed the changes in the oxidation state of the bulk and surface species on the Co-OH nanoflakes supported on the KMO. XPS and Time of flight secondary ion mass spectrometry (ToF-SIMS) analysis demonstrated that the layer of CoOxHy deposited on the KMO and its electronic structure strongly depends on the anion of the precursor used. In particular, it was found that Cl- favors the formation of Co-OH, changing the rate-determining step of the reaction, which enhances the catalytic activity towards the OER, producing the most active OER catalyst in alkaline media.

In this thesis, a Discrete Multi-Physics model based on Smoothed Particle Hydrodynamics is developed to simulate a Rayleigh collapse of a single bubble. All the simulations were run on a modified version of the open source software LAMMPS and visualised on OVITO. Initially a 2D model is validated by simulating a phenomenon that shares many similarities with a collapse mechanism, the interaction of a shock wave with a discrete gas inhomogeneity, showing similar performance to classic mesh based CFD. The model is then used to simulate a 2D Rayleigh collapse and validated against the 2D Rayleigh-Plesset equation for both empty and gas filled cavity. The validated model is used to investigate the role of heat diffusion at the gas-liquid interface of the cavity, and to study non-symmetrical collapse induced by the presence of a nearby surface. Enabling heat diffusion at the gas-liquid interface allowed to identify five different possible behaviours that range from isothermal to adiabatic, while the results of non symmetric collapse show that the surface is hit by a stronger shock when distance between the center of the cavity and the surface is zero while showing more complex double peaks behaviour for other distances. In the final chapter a 3D model is used to model an attached non-symmetrical collapse and its hydrodynamic is compared with the equivalent 2D case.

The generation of high-affinity antibodies is critical for natural and induced protective immune responses and production of efficacious therapeutic monoclonal antibodies. Germinal centre matured plasma cells (GCmat PCs) produce these long-term high-quality responses. The signals and mechanisms by which GCmat PCs are induced are not fully elucidated. We developed parallel strategies to generate novel gene manipulated mice where GCmat PCs are easily identifiable by fluorescent markers controlled by GC-specific S1pr2 and PC-specific Prdm1 expression. Combination of validated transgenic alleles generated the High-Affinity LOw-affinity (HALO) PC mouse, able to distinguish between GCmat PCs and extrafollicular PCs. Kinetic assessment of early GC output confirmed previous results from our lab that GCmat PC output peaks early then declines. Furthermore, we detected transient re-expression of CD38 in GCmat PC precursors and GCmat transitioning PCs, an intriguing result with a yet undetermined function. In tandem, we developed and characterised a novel split fluorescent reporter, diSplit670. CRISPR/Cas9 enhanced targeted insertion placed each part of diSplit670 under control of endogenous Prdm1 and S1pr2 in separate mouse embryonic stem cell (ESC) lines. To-date, germline transmission has been achieved for one targeted ESC line. These novel tools would be uniquely suited to interrogate the biology of GCmat PCs.

Glycoprotein non-metastatic melanoma protein B (GPNMB) is a transmembrane protein highly expressed in multiple tumours, including classical Hodgkin lymphoma (cHL); a tumour in which the malignant Hodgkin and Reed-Sternberg (HRS) cells rely on the tumour microenvironment (TME) to survive and evade detection by the host immune system. In this study, I explore the contribution of GPNMB to immune evasion of cHL as a novel therapeutic target. I have shown that GPNMB is highly expressed in tumour-associated macrophages (TAM) in cHL tissues but expression is highly variable. In vitro differentiation of macrophages (from CD14\(^+\) monocytes) was shown to be associated with an increase in GPNMB expression, including the generation of a soluble form of GPNMB (sGPNMB). sGPNMB partially inhibited T-cell activation and T-cell recognition of Epstein-Barr virus (EBV)-specific epitopes in cHL cell lines. Preliminary experiments indicated that the inhibition of T cell activation by GPNMB could be overcome with GPNMB neutralising antibodies. This work provides evidence in support of the hypothesis that GPNMB can mediate an immune checkpoint that inhibits anti-tumour cytotoxic T-lymphocytes (CTL). It provides a basis for further investigations designed to explore the therapeutic potential of targeting GPNMB in cHL.

Hevea brasiliensis, the Par`a rubber tree, is the primary source of natural rubber (NR). NR is harvested through the wounding, or tapping, of rubber trees, and the collection of the latex that flows from the wound. Serial tapping and ethylene stimulation can boost the yield of latex harvested after each tapping, though the mechanisms that underpin this increase in yield have not been fully characterised. This study has generated whole-genome transcriptome and methylome datasets, in order to explore the regulatory mechanisms that underpin latex production in response to tapping and ethylene stimulation. Analyses show key patterns of differential expression across genes of functional significance, including a large number of genes under the control of endogenous-ethylene, whose transcriptional response to tapping is enhanced by exogenous ethylene. This includes numerous DNA methylation relevant genes, indicating an epigenetic role in latex production, and therefore prompting the generation of bisulfite-seq libraries. Collectively, these experiments will provide great insight into the mechanisms underpinning latex production.

The main aim of this study was to gain a deeper knowledge of flow behaviour and scour formation adjacent to bridge piers and offshore wind turbine foundations in shallow water using numerical modelling. Based on the literature, the capability of the previous models in the prediction of scour around the piers was attached to the pier shapes and its orientation angle to the flow. Hence, in this study, a coupled approach between a hydrodynamic and a morphodynamic model was implemented. The model was validated against eight different test cases for flow and sediment transport modelling. Overall, the results confirmed that the developed model has an acceptable capability in predicting local scour independent of the pier cross-section shapes. Furthermore, the model was applied to predict the flow field, bed shear stress and the development of local scour around a square-shaped pier with 45˚ (diamond) and 90˚ (square) orientation angle to the flow, a hexagon-shaped pier with 30˚ and 60˚ orientation angle to the flow and a group of two identical side-by-side circular cylinders with different spacing ratios G⁄D= 1, 2 and 3 under current flow and live-bed condition, where D represents the pile diameter and G is the distance between the piles. In the case of modelling a single pier, it was found that the maximum and minimum normalized quasi-equilibrium local scour depth occur for the square-shaped piers with 90^° and 45^° orientation angles to the flow, respectively. The normalized scour depth in quasi-equilibrium condition for the square pier was found to be almost twice (1.94) of that for the diamond pier. This ratio in the case of the hexagon-shaped pier for 60^° and 30^° orientation angles to the flow decreased to a value of 1.09. From these results, it was realized that an increase in the number of sides of the piers cross-section decrease the impact of the change in orientation angle to the flow in terms of normalized scour depth. It was also inferred that with an increase in the number of sides of the pier's cross-section the resultant normalized scour depth in equilibrium condition tend to get closer to that of the circular pier regardless of pier orientation angle to the flow. In the case of the local scour simulation around a group of two identical side-by-side piles, it was noticed that a deeper scour hole was observed in the area between the piles in all cases. However, due to the stronger turbulent structures and higher contraction of streamlines in the case of G⁄D=1, a deeper hole was observed adjacent to the piles in comparison to that of two other cases. The effect mitigated with an increase in spacing between the piles. The results showed that for G⁄D≥ 3 the pile group effect gradually vanishes, while it is assumed that the pile group effects are completely negligible for G⁄D≥ 5. An equation was fitted for the computed equilibrium scour depths for different spacing ratios which can be used for the prediction of the maximum scour depth adjacent to two side-by-side piles under steady current flow and live-bed condition for a specific range (G⁄D≥ 1).

This thesis consists of two volumes submitted towards the Doctorate in Forensic Clinical Psychology. Volume I comprises three research chapters. The first chapter presents a systematic review of the literature exploring meditative approaches in correctional settings. A number of meditation and mindfulness approaches are discussed in relation to psychological and physical wellbeing, criminogenic needs, and recidivism in offender populations. The second chapter is an empirical research study which sought to validate a measure of a Buddhist derived concept of mindfulness, insight, in a UK undergraduate university sample. The third chapter is a public dissemination document which provides a brief and accessible overview of the review and empirical components. Volume II comprises five forensic clinical practice reports. The first presents the case of Paul, a 22-year-old male presenting with verbal and physical aggression, formulated from two psychological models. The second presents a service evaluation of the involvement of service users in the care planning process in a secure inpatient hospital. The third presents a single-case experimental design of the case of Jane1, a 22-year-old female with a mild learning disability and a suspected autistic spectrum condition, presenting with anxiety relating to pregnancy and motherhood. The fourth presents the case of Claire1, a 17-year-old female experiencing symptoms of low mood, anger, and difficulties in her relationship with her mother. The fifth presents an abstract of an oral presentation about the case of John1, a 34-year-old male serving an IPP sentence in a Category B prison, presenting with frequent self-harming behaviour.

Oxidative stress and subsequent oxidative DNA damage play an important role in the process of ageing and are also linked to numerous different diseases. A prevalent type of oxidative DNA damage is the oxidation of guanine, generating 8-oxoguanine. It is therefore unsurprising that 8-oxoguanine has been found to be elevated in the genome in a range of pathological conditions, including multiple cancers. However, whether 8-oxoguanine drives the progression of these diseases or is a result of the disease process itself is often unclear. 8-oxoguanine is recognised in genomic DNA by the glycosylase, 8-oxoguanine glycosylase-1 (OGG1), which initiates removal by the base excision repair pathway. To study cellular function, OGG1-null mouse models and in vitro cell models have been previously generated. Primarily, this thesis is an investigation into the cellular role of OGG1 and functional changes in cells that lack OGG1-activity. We show that OGG1-null cells exist in a state of oxidative stress, possessing a persistent elevation of intracellular reactive oxygen species. Further increasing the oxidative stress burden by the depletion of the antioxidant, glutathione, induces apoptosis in OGG1-null cells. Surprisingly, however we show that induction of apoptosis is not accompanied by an accumulation of genomic 8-oxoguanine. This finding highlights the importance of backup repair mechanism, which are able to repair 8-oxoguanine in the absence of OGG1. Using gene expression and biochemical studies, we have identified two potential backup mechanisms: components of the nucleotide excision repair pathway (ERCC1, ERCC4 and ERCC5) and the Nei-like DNA glycosylase-3 (NEIL3). Lastly, we have confirmed that OGG1 possesses an additional cellular role, acting as a signalling molecule to activate small GTPases, activating multiple downstream pathways. This stresses the importance of not only considering OGG1 as a glycosylase, with multiple additional functions having already been investigated.

Maternal mental health conditions have been shown to be among the most disruptive influences on the mother-infant bond and their reciprocal social relationship, often resulting in long-lasting, often adverse infant outcomes. However, the mechanisms underlying this are poorly understood. Reduced placental expression of the imprinted gene PEG3 has been linked to prenatal depression in human pregnancy. In mice, intrinsic loss of function of Peg3 in dams has previously been linked to poor maternal care. Reduced expression of Peg3 in the murine feto-placental unit has also been shown to indirectly impair maternal care and increase anxious-like behaviour in wild-type dams. It has been hypothesised that this indirect effect on maternal behaviour may result from defects in the development of the placental endocrine lineages. Together, these studies suggested that reduced placental PEG3 expression may play a causal role in driving maternal mood symptomology and adverse infant outcomes in human pregnancy. Using a Peg3 loss-of-function mouse model, this thesis replicated previous work linking reduced offspring Peg3 expression to impaired maternal care and increased anxiety-like behaviours in wild-type dams with 100% mutant litters. For the first time, this thesis extended previous work, characterising the behaviour of wild-type dams with litters composed of both mutant and wild type pups. Behavioural characterisation of Peg3 mutant offspring revealed a sexually dimorphic effect of Peg3 disruption, highlighting a male-specific deficit in neonatal social behaviour in mutant mice. Later life social impairments were also observed in both mutant and wild-type littermates. Inclusion of wildtype controls allowed for these behavioural changes to be attributed to the adverse maternal environment, likely induced by placental endocrine sufficiency, as opposed to intrinsic loss of function of the Peg3 gene. PEG3 expression and social communication were also explored in mother-infant dyads from the Grown in Wales human cohort. There were no associations with PEG3 and mother-infant social behaviour. However, sexually dimorphic outcomes in response to prenatal depression were observed. Specifically, prenatal maternal mood symptomology was associated with negative aspects of vocalisation in mothers with female, but not male infants. In summary, this thesis provides support for the relationship between placental endocrine dysfunction, maternal behaviour, and offspring outcomes in a Peg3 loss of function model. It also contributes to a growing body of literature suggesting that males are differently affected by disruption of Peg3 expression. Finally, though no association with PEG3 was reported in the human cohort, this thesis highlights the need to study sexually dimorphic outcomes in relation to perinatal maternal mood symptomology, as a mother's behaviour may be influenced by the sex of their infant.

This PhD thesis identifies, examines and characterises all parts of the carbon capture and utilisation (CCU) value chain to assess its potential to be used as a CO2 emission mitigation option and, by extension, support global warming mitigation efforts. It puts CCU value chains in the context of global warming mitigation by establishing the relationship between increasing CO2 emissions, global warming and the efforts of the European Union to reduce them. CCU value chains can be defined as the process of generating profit by capturing CO2 from CO2-emitting sources and transporting it to CO2 receivers for utilisation, within a process or the production of commercial products. The examination of CCU value chains starts by identifying the chain's individual steps followed by a breakdown of the steps in four sections that make up the chain: utilisation, CO2 sources, CO2 capture and CO2 transportation. Each section is introduced by discussing its purpose, process and limitations within the chain, followed by a characterisation of the key elements and identifying and specifying factors that could be used in optimisation. Utilisation happens at CO2 receivers where CO2 is sold for profit and utilised in a process or for the production of commercial products; CO2 sources produce the CO2 emissions; CO2 capture technologies capture CO2 at the source and prepare it for transportation; and CO2 transportation is responsible for delivering CO2 at the receiver for utilisation. Numerous attempts have been made in the past aimed at optimising individual steps of the chain, and various attempts have been made to integrate approaches and achieve optimisation for more than one of the steps, but these have proved difficult because of their high complexity. This PhD thesis chose a different and simpler approach that requires fewer variables and provides a quick and reliable holistic solution to propose optimised regional CCU value chain schemes by integrating all steps of the CCU value chain. It was concluded that CCU value chains show high potential for CO2 emission mitigation if they are rigorously assessed, tailored and optimised for a specified region before application. The novel contributions of the thesis are: • Knowledge base of CO2 receivers and their characterization • Knowledge base of CO2 sources and their characterization • Framework for the matching of CO2 sources with CO2 capture technologies based on their compatibility • Models for estimating CO2 capture cost • Algorithm and business model for CCU value chain optimisation The first four concepts have been integrated within an algorithm, the main novelty of the thesis, which optimizes the implementation of CCU value chains, and a business model that proposes realistic CCU schemes in a given region. The developed algorithm and business model can map the CO2 sources and receivers in a specified region and select the sets of optimal solutions based on the optimisation preference of the user for the development of CCU value chains by matching CO2 sources and receivers based on (i) the technological compatibility and maturity of technologies, (ii) CO2 capture costs, (iii) CO2 transportation costs, (iv) CO2 utilisation costs and (v) profit within a defined project lifetime. The algorithm and the business model have been validated using real life examples (industry level, regional level, and national level) and have been also implemented in an online platform for enabling symbiotic value chains for solid waste management (the development of which happened as part of the Interreg V-B "Balkan Mediterranean 2014-2020" SWAN project where we as part of the University of Huddersfield were invited as experts in industrial symbiosis field, but which was outside the scope of this thesis). The application and validation of the algorithm and business model in these areas demonstrate the strengths of the novel concepts and their potential to commercialise CCU value chains and contribute to global warming mitigation, and also to be used in areas other than CCU value chain optimisation.

Surface reconstruction, or reverse engineering of a surface, is the practice of combining data collection and analysis techniques for retrieving measurement data from an object with the aim of generating a digital representation of the object, where the original is either missing or incomplete. Recovering the digital representation of a scanned physical shape will, in most cases, be in the form of a three-dimensional (3D) point cloud. However, the process of capturing the data will inevitably lead to a point cloud contaminated by the uncertainty of measurement and imperfections in the measured object. To reverse engineer an object, especially a used one, consideration must be made for any manufacturing defects (tolerances or imperfections) and in-service damage. Given the significant challenges in designing a surface reconstruction algorithm, this project aims to develop a framework for reverse engineering/ performing surface reconstruction for damaged objects and develop a method for identifying and localising damage on the surface of a model with an accuracy of a few tens of micrometres. This accuracy level is significant due to the resolution of the generated data. The workflow for reverse engineering considered in this thesis is capturing point cloud data (PCD) using a laser scanner mounted on an Articulated Arm Coordinate Measuring Machine (AACMM). Preprocessing of the raw points is performed to mitigate noise, sparse data, and contamination from the measuring system and identify properties that influence the data file structure, such as interoperability. Before performing detection, it is necessary to identify geometric features from the PCD, since an artefact's "nominal" design geometries are unknown. Hence, a linear fitting method is applied to the input point cloud data, and the system is optimised to obtain a good level of robustness of fitting the line of best fit for estimating "nominals". A combination of edge detection, profile monitoring, model sectioning, and machine learning are used to identify potential damage locations on a model's surface. The processes are ideally designed to be reproducible, so with reduced human influence. Slices produced from model sectioning of the PCD are unwrapped to create planar (X-Y) data at the micrometre-level, suitable for using machine learning (ML). The ML is trained and tested to classify the spatial-sequences of the data as either "good" or" damaged". When applied to a potentially damaged surface, the sequence classification can then identify the location of damage along the height axis of a model. Future work would generalise this into three dimensions. The performance of the proposed system is evaluated through mathematical simulation of noisy point cloud data on objects with simulated defects and by physical validation using measurement data. The results indicate that the method can detect and localise areas of damage. This is a stepping-stone to automated repair of damaged surfaces by adaptively reproducing a copy of the original part, considering deviation of the "as manufactured" from the design intent. The thesis contribution to knowledge is a novel method of damage detection and localisation at the micrometre level as a step towards implementing a surface reconstruction framework, which has also been developed in this study.

The emergence of life on Earth developed a system of DNA containing the information inherited in the reproduction of organisms, directing also the variability of life as this changes. The language using for deciphering the DNA is the genetic code, which is degenerated as 20 amino acids are codified by 64 codons, or triplets of nitrogenous bases (adenine, guanine, cytosine and thymine) in the nucleotides that form the DNA. 59 codons have other/s synonymous codons, and those are not used randomly but are a product of different phenomena. The drivers of codon usage bias can be mutations, genetic drift and natural selection processes. The study of these codons can provide evolutionary information, like gene family evolution and organisation, can offer an insight in population size, and also can be used to create software to predict ORFs and coding regions. Codon study also can be important in biotechnology for heterologous expression, as the codon used have a control over the efficiency and accuracy during translation, resulting in determinant for protein folding, growth rate and avoidance of defective proteins. Opisthokonta is a group of organisms, close to Apusomonadida, and gathering Holozoa (Ichthyosporea, Filasterea, Choanoflagellatea and Metazoa) and Holomycota. The present study uses the transcriptome of 20 choanoflagellates species, Sphaeroforma arctica (Class Ichthyosporea), Thecamonas trahens (class Apusomonadea) and five early branching metazoans (Amphimedon queenslandica, Mnemiopsis leidyi, Nematostella vectensis, Thelohanellus kitauei and Trichoplax adhaerens) to study the codon usage bias and anticodon related phenomena in the mentioned species. Analysis revealed optimal codons for each organism, with a GC ended codons enrichment in T. trahens, unicellular Opisthokonta, and retained after the transition to pluricellularity (with divergences in M. leidyi and N. vectensis). All the unicellular species showed evidence for natural selection upon translational accuracy and efficiency, even the less biased ones. In metazoan only translational efficiency was evidenced, with lower strength of selection, and poor support for translational accuracy in ctenophoran and cnidarian species. For other strategies giving advantages on protein translation, indirect evidence of deamination in high degenerated amino acids was found in all the species studied, while direct evidence was found in S. arctica, A. queenslandica, M. leidyi and N. vectensis. Additionally, tRNA presenting isotype-anticodon disagreements were found in apusomonadean, ichthyosporean, metazoan species and in S. rosetta. Deamination, and possibly chimeric tRNA appeared early in evolution, before the divergence of Apusomonadea and Opisthokonta, benefiting the use of optimal codons through evolution until the apparition of the latest metazoan organisms.

Autophagy is a vital homeostatic pathway essential for cellular survival and human health. It primarily functions as an intracellular degradation process for the turnover of unwanted cytoplasmic materials such as aggregation-prone proteins, damaged organelles and invading pathogens. Defective autophagy contributes to the pathology of a myriad human diseases including neurodegeneration and infection, and thus therapeutic exploitation of autophagy presents an attractive opportunity. The work described in this thesis involves establishment of human embryonic stem cell (hESC) models of autophagy for studying its homeostatic role and developing screening platforms in physiologically-relevant human cell-types of interest. The autophagy-deficient (ATG5-/-) hESC-based system was used to address how loss of autophagy leads to cytotoxicity; an important question relevant to neurodegeneration. The ATG5-/- hESCs and hESC-derived neurons exhibited basal cytotoxicity accompanied by a specific metabolic defect. Depletion of NADH was found to be mediating cytotoxicity during autophagy deficiency via mitochondrial depolarisation, since boosting cellular NADH levels improved cell viability and mitochondrial membrane potential in ATG5-/- neurons. This intervention was also cytoprotective in a human induced pluripotent stem cell (hiPSC)-based disease model of a neurodegenerative lysosomal storage disorder associated with autophagy dysfunction, and thus could be relevant for related neurodegenerative conditions. The autophagy reporter (mCherry-EGFP-LC3) hESC-based system was developed for identifying potent autophagy inducers enhancing pathogen clearance in macrophages; relevant for combating infection. The hESC-derived macrophages were characterised to confirm their cellular identity. These isogenic platforms can be of basic and biomedical relevance for mechanistic studies of autophagy regulation including human cell-type specific effects, and drug discovery of autophagy modulators for therapeutic applications in human diseases.

This PhD deals with the development of Microtubular Solid Oxide Fuel Cells (µT-SOFC) at the cell and sub-stack level. The study provides a body of work as a basis for future development in the field. The information has been acquired and interpreted empirically and numerically. µT-SOFCs are suited to a broad range of applications with power demands ranging from a few watts to several hundred watts. µT-SOFCs possess inherently favourable characteristics over alternate configurations such as high thermo-mechanical stability, high volumetric power density and rapid start-up times, lending them for particular value use in portable applications. Efficient current collection and interconnection constitute a bottleneck in the progression of the technology and addressing this was the focus of this study. An empirical study to determine the effect of varying current collector position and number was conducted. A mathematical model with its underpinnings in a resistance path model was constructed to interpret the empirical findings. A multiphysics CFD model mirroring the experimental setup was constructed to build on the insight provided by the mathematical model. The model was validated and fitted with experimental data and was reliable for prediction of µT-SOFC performance within set parameters. Contacting of each electrode is most simply and most typically achieved from the cell exterior at the expense of available active cell area and durability. Building on the knowledge gained in the empirical and numerical studies, a novel method of internal current collection was proposed; collecting current from multiple points along the inner wall of the anode supported cell. The current collector also acted as a flow turbuliser, enhancing the flow and thermal distribution within the fuel cell. To reduce the contact resistance of the internal current collector, a contacting technique was identified, namely brazing. Brazing enabled the formation of multiple, physical joints along the interior of the µT-SOFC. The brazing improved the current collector performance and durability. The study featured the development of a novel braze and braze application technique. CAD and CFD were used for the optimisation of a 4-cell µT-SOFC manifold, stack, and operating conditions. The stack design used data from the experimental testing of cells with the novel, brazed internal current collector. Two 4-cell stacks with two current collector variations, both using the novel, brazed internal current collector design, were fabricated and electrochemically tested. Data from performance and durability testing of the stacks was presented and the best design identified.

Certificate transparency is a promising solution to publicly auditing Internet certificates. However, there is the potential of split-world attacks, where users are directed to fake versions of the log where they may accept fraudulent certificates. To ensure users are seeing the same version of a log, gossip protocols have been designed where users share and verify log-generated data. This thesis proposes a methodology of evaluating such protocols using probabilistic model checking, a collection of techniques for formally verifying properties of stochastic systems. It also describes the approach to modelling and verifying the protocols and analysing several aspects, including the success rate of detecting inconsistencies in gossip messages and its efficiency in terms of bandwidth. This thesis also compares different protocol variants and suggests ways to augment the protocol to improve performances, using model checking to verify the claims. To address uncertainty and unscalability issues within the models, this thesis shows how to transform models by allowing the probability of certain events to lie within a range of values, and abstract them to make the verification process more efficient. Lastly, by parameterising the models, this thesis shows how to search possible model configurations to find the worst-case behaviour for certain formal properties.

With the continuous and rapid development of the economy and the acceleration of urbanisation, public transport in cities has entered a period of rapid development. Urban rail transit is characterised by high speed, large traffic volume, safety, reliability and punctuality, which are incomparable with those of other forms of public transport. The traction power supply system (TPSS) is an important part of an electrified railway, and its safety issues are increasingly prominent. Different from the substation in a general power system, the load of a TPSS has a great impact on the traction transformer; moreover, in order to ensure normal operation of the train in case of failure, the traction substation must be able to access a cross-district power supply, as it has a high demand for reliable operation. The safe and reliable operation of DC TPSSs is the basis of the whole urban railway transit system. Previous studies have investigated the reliability of the TPSS main electrical wiring system. However, the impact of traction load and the actual operation of trains on system reliability has not been considered when designing a DC railway power supply system. The purpose of the research for this thesis is to find an optimal system operation plan for urban railways, considering load characteristics. This thesis begins with a review of the main arrangements of DC railway power supply systems and the literature on railway reliability studies. A model of single train simulation and a power supply system is established in MATLAB. The developed simulator is then integrated with a TPSS reliability model to evaluate the energy and reliability performance of DC railway power systems. Based on the train traction load model and train schedule, a comprehensive method for evaluating a DC TPSS considering traction load is proposed. Through simulation of the actual operation of the train group, the system energy consumption and substation life loss generated under different train operation diagrams and schedules are compared to provide a reference for the reasonable design of the timetable. Taking the life loss and energy consumption of the whole TPSS as the objective function, a genetic algorithm is used to optimise the train speed, coasting velocity, station dwell time and headway to find the optimal operation strategy. This is illustrated with a case study of the Singapore East-West metro line. The study has addressed the following issues: development of a multi-train power simulator, evaluation of reliability performance, and finally the search for an optimal operation plan. The train running diagram and timetable are optimised jointly. This can help railway operators make decisions for an optimal operation plan and reduce the operation risk of the power system.

A comprehensive understanding of the steam-side oxidation and spallation properties of austenitic stainless steels is required to enable the use of these materials in ultra-supercritical plant in order to raise efficiency. Isothermal oxidation of austenitic steels Super 304H, 347H FG and HR3C was performed in laboratory air and deoxygenated flowing steam at atmospheric pressure for oxidation periods of up to 3000 hours and 1300 hours respectively. The oxides on two inner tube surface finishes were characterised - A pickled surface finish, as would be installed in plant (Super 304H, 347H FG, HR3C), and a shot peened surface finish (Super 304H). Oxidation in flowing deoxygenated steam was found to increase the oxidation rate of austenitic stainless steels compared with oxidation in laboratory air. Pickled Super 304H and 347H FG "18-8" austenitic steels exhibited larger mass gain and oxide thicknesses than 25Cr HR3C in all atmospheres. The inner surface of shot peened Super 304H exhibited dramatically thinner oxides in both air and steam. The pickled austenitic stainless steels formed a duplex oxide featuring an outer iron rich oxide and inner spinel oxide containing the alloy constituent elements. The outer iron oxide grown in laboratory air was identified as haematite and that in deoxygenated steam, magnetite. At 650 and 700 °C in steam, a protective Cr-Mn spinel formed at the base of the spinel oxide. The parabolic rate constant of this oxide was found to be an order of magnitude larger than that of pure chromia, and hence less protective. The oxidation kinetics of the spinel and outer oxide layers at 650 and 700 °C were found to follow two regimes throughout oxidation. Initially growth of these layers follows an accelerated parabolic regime, until the complete formation of a protective Cr-Mn spinel at the base of the inner spinel oxide. Following this, the growth rate is dramatically reduced. This is caused by reduced cation diffusion through the Cr-Mn layer. Spallation of the outer oxide of a number of samples occurred following oxidation in ineffectively deoxygenated steam. The presence of 20% or more haematite in the outer oxide was a prerequisite for spallation. The spallation interface was often within the magnetite layer and, occasionally, at the spinel/outer oxide interface. The study was extended to include oxidation tests in flowing air saturated steam. These were performed to promote spallation of the inner surface oxide on 347H FG. The strain energies within spalled oxide were calculated and compared to those found in the literature. A value of γ' = 13 Jm-2 was calculated for the effective fracture energy for the outer oxide on 347H FG, in good agreement with literature values for type-316 austenitic steel. Analysis of the oxide scales revealed that oxide buckling was the most common mode of failure, and that spallation of these oxides satisfies the critical strain energy criterion. Spallation maps for haematite and magnetite growth on 347H FG were constructed, and data was fit, and corresponded to the area at which unstable buckling or wedging was expected, in agreement with the experimental data. An additional feature of oxidation in air saturated steam was the formation of oxide blisters on the inner surface of 347H FG. These blisters are proposed to nucleate and grow during oxidation, and rupture on cooling.

The aim of the present thesis was to extend existing research on authentic leadership in sport and develop a better understanding of the impact of authentic leadership on athletes, through investigating the relationship between coaches' authentic leadership and a range of reported athlete outcomes. Authentic leadership is a style of leadership which has received little attention in sport. However, due to its proposed core components and focus on followers' development and creating trusting relationships with followers, it could prove promising in promoting positive athlete outcomes. After reviewing the literature in Chapter 1, Chapter 2 presents Study 1 which examined how authentic leadership is unique when compared to the dominant sport leadership theory i.e., transformational leadership. Specifically, Study 1 examined whether authentic leadership is empirically distinct from transformational leadership and what it adds to transformational leadership in terms of predicting athletes' commitment and enjoyment, when controlling for transformational leadership. Authentic leadership was found to be positively correlated to transformational leadership, suggesting evidence of overlap between the leadership styles. However structural equation modelling revealed authentic leadership to also be different to transformational leadership and demonstrated that it predicted athlete enjoyment and commitment beyond that explained by transformational leadership. After finding evidence for the divergent validity of authentic leadership compared to the prominent leadership theory in sport, Chapter 3 consists of a longitudinal study which examined whether authentic leadership would be directly related to changes in athletes' enjoyment, commitment, and prosocial behaviours, or indirectly related through changes in trust and cohesion over time. Authentic leadership was found to be directly related to changes in athletes' prosocial behaviours and enjoyment over time. Authentic leadership was also indirectly related to athletes' enjoyment via changes in cohesion over time, and indirectly related to prosocial behaviours via changes in trust and cohesion over time. Chapter 4 consists of an experimental study which examined the effects of authentic leadership on athletes' trust, achievement, and moral outcomes. The results of Study 3 demonstrated that athletes in the high authentic condition reported higher trust, commitment, enjoyment, anticipated guilt and lower aggression compared to participants in the low authentic leadership and neutral conditions. Chapter 5 consists of Study 4 which examined the potential effectiveness of an intervention designed to increase coaches' demonstration of authentic leadership. This authentic coaching intervention was pilot tested and evaluated by examining its efficacy on athletes' perceptions of their coaches use of authentic leadership and a range of reported athlete outcomes. Study 4 indicated that athletes in the intervention group reported significantly higher perceptions of authentic leadership, as well as higher enjoyment and prosocial behaviours compared to the control group. Overall, the findings of the thesis suggest that coaches should be encouraged to demonstrate behaviours indicative of authentic leadership as this was found to have positive consequences for their respective athletes and teams.

Behavioural scales are the clinical standard to assess consciousness in patients diagnosed with disorders of consciousness (DOC), although patients remain unresponsive to environmental stimuli beyond reflexes. Electrophysiological studies have identified modulation of brain activity in some patients when following commands; however, command-following requires complex brain processing that most patients lack due to the severity of their brain injury. Passive EEG paradigms differentiating automatic from strategic processing have detected residual cognitive processing in DOC patients. The local-global paradigm measures trial by trial violations of local expectations (short time-scales), which elicit a mismatch negativity (MMN) response that reflects automatic processing; whereas violations of global expectations (longer time-scales), that are generated using contextual information, show a positive ERP component (P3b) that accounts for controlled processing. Following the same rationale, this thesis includes four experiments that investigate the influence of semantic local and global expectations on word processing, aiming to assess residual language comprehension in DOC patients. By employing a relatedness proportion paradigm in a semantic priming task, we provide a 'local' context within trials (i.e., related/unrelated word-pairs), as we simultaneously manipulate global expectations by providing a context across the task (i.e., prime validity: cueing participants about the probability of a related target following the prime). We first report three behavioural visual studies in healthy participants, which suggest that individuals show greater priming effects in high validity contexts relative to low validity contexts (i.e., relatedness proportion effects), as individuals use the global context strategically to generate expectations about the target. Strategic involvement is supported by self-report measures, as only individuals that applied conscious strategies while performing the task showed these behavioural global effects. In a subsequent study we investigate the neural correlates of the generation of local and global expectations in this paradigm. The results show an earlier 'local' prediction error ERP effect around 250ms; followed by a later 'global' effect approximately at 350ms that interacted with the global context. We then adapt the task for DOC patients to identify auditory markers of strategic expectancy generation in this paradigm. The results in the healthy group suggest an early 'local' ERP effect (around 350ms), showing a prediction error ERP signal; and a later effect (around 550ms) reflecting violations of the 'global' semantic context i.e., larger error signal for unexpected targets in high validity relative to low validity context. Moreover, we present two DOC patient cases, where we detect a 'local' ERP effect in a patient (MCS diagnosis), whereas no ERP effect is observed in the other case (VS/UWS diagnosis). In conclusion, we propose this auditory ERP paradigm as a tool to detect residual language comprehension in DOC patients, as it detects hierarchical differentiated effects for strategic and non-strategic semantic expectations, similar to the local-global paradigm but in the clinically relevant domain of language processing.

Biological cell injection is an effective technique in which a foreign material is directly introduced into the target cell. Intracytoplasmic Sperm Injection (ICSI) is a microinjection technique which is used for infertility treatment. In this technique, a single sperm cell is directly injected into an oocyte using micropipettes. The operations in this application are manually controlled by an embryologist and more importantly, this reduces the accuracy, repeatability, and consistency of the operation. Therefore, the full automation is a prerequisite for microinjection operations, particularly in ICSI application. This thesis focuses on enhancing the microinjection procedure by developing vision-guided processes prior to and during the operation. Initially, a vision-controlled technique was proposed to align the injection and holding pipettes in three orthogonal axes which is essential for successful microinjection. To conduct reliable injection, the vibrational displacement of the injection pipette's tip needs to be evaluated and improved before the operations continue further. A novel vision-based sensor was developed to measure the displacement changes at the tip in three orthogonal axes. By employing the developed vision sensor, the effect of injection speed on vibrational displacement creation was analysed to determine the value of various injection parameters, such as force fluctuation, and penetration force on cell damages. An ultimate automation task is required in microinjection to position the randomly located biological cell within the Petri dish to the system's field of view. The proposed technique fills a gap in the literature by proposing a real-time cell recognising and positioning system that can be employed with different types of biological cells at various maturation stages, as well as with different microscope types that are being used in microinjection applications.

Double strand breaks are highly deleterious lesions, and their inefficient repair can lead to mutations and chromosomal translocations. These events cause severe phenotypes such as cancer, immunodeficiencies and neurodegenerative disease. Double strand breaks can be repaired by several mechanisms and the decision of repair pathway is crucial to faithful repair and preservation of genomic integrity. Repair pathway choice is regulated by numerous factors including cell cycle phase, kinetics of repair and resection of DNA at the break site. Recent studies applying fluorescence microscopy have provided compelling evidence that repair factors are assembled at the break site in a highly spatially and temporally organised manner, and that the spatial dimension of repair is critical to faithful damage resolution. The aim of this thesis is to characterise the spatial organisation of double strand break repair factors to corroborate a spatial model of repair progression. In this work, I applied novel fluorescence microscopy technique, Expansion Microscopy. This technique enables three-dimensional nanoscale imaging of specimens on a conventional microscope. This is achieved by a specialised chemical preparation in which cells are embedded into a polyelectrolyte gel which can subsequently be swollen by the addition of water. The net effect of this is a de-crowding of molecular information in the specimens and lateral and axial resolution of 70 nm and 200 nm, respectively. Super resolution images can be acquired rapidly, with minimal constraints on fluorophore choice, and without need for sophisticated hardware or software. Specimens are optically cleared during the preparation, which removes the effects of light scattering through the sample. Increased imaging depths and volumes are achievable as compared to other super resolution imaging modalities. In this work, I demonstrate an ability to visualise thousands of repair structures using expansion microscopy which are analysed by a spot detection-based analysis and provide a novel quantitative insight into spatial organisation of repair foci during double strand break repair.

Liver transplantation is a successful treatment for acute and chronic liver failure, yet the field faces challenges in the forms of deteriorating graft quality, increasing demand and more complex logistics. Normothermic machine perfusion is able to preserve donor livers for up to 24 hours and liver function can be objectively assessed during the perfusion process. Criteria that aim to establish donor liver viability have been developed through the perfusion of livers discarded for clinical use. The criteria are based upon metabolic and physiological parameters and have been tested within a clinical pilot study of five transplants using discarded organs. All grafts functioned immediately and all patients survived. Following on from this success, VITTAL (Viability testing and transplantation of marginal livers) - a Wellcome Trust funded trial - was designed and carried out. This saw the transplantation of 22 patients following 31 perfusions of discarded donor livers. The results validate the viability criteria and demonstrate that end-ischemic normothermic machine perfusion enables the safe transplantation of a significant proportion of currently unutilised livers and is associated with increased graft utilisation, extended preservation time and improved logistics. The use of Hemopure, a haemoglobin-based oxygen carrier, was also investigated and has been shown to have logistical, rheological and immunological advantages over packed red cells when used in the perfusion fluid. The final part of this thesis explored the use of machine perfusion devices to deliver cellular therapy to marginal donor livers. The results demonstrate the technique is feasible, with multipotent adult progenitor cells being delivered directly into the target organ allowing them to secrete a host of soluble factors that are known to have anti-inflammatory and immunomodulatory effects.

The scientific community has long been intrigued by the potential use of human waste odour in medical diagnosis and disease monitoring. Human waste odour analysis offers a fast and non-invasive way of testing and has the potential as an early disease diagnostic method. Cancer (particularly pancreatic and colorectal cancer), which has been known to be one of the deadly diseases of all time due to the late prognosis, may benefit from odour analysis as the current diagnostic method using highly invasive procedures such as colonoscopy. Other diseases, such as inflammatory bowel disease (IBD), which also employed colonoscopy in its diagnostic procedure, could also benefit from the application of human waste-based investigation. In this work, data analysis pipelines employing five machine learning algorithms were developed for analytical instrument electronic nose (E-nose), field asymmetric ion mobility spectrometry (FAIMS), gas chromatography ion mobility spectrometry (GC-IMS), and gas chromatography time of flight mass spectrometry (GC-TOF-MS). These methods were used in 8 studies to investigate the potential of non-invasive diagnosis and monitoring of pancreatic cancer (PDAC), colorectal cancer (CRC), and IBD through urine, breath, and faecal odour analysis. The PDAC studies included 285 subjects: 126 with PDAC, 45 with Chronic Pancreatitis (CP), and 114 controls, and utilised FAIMS, GC-IMS, and GC-TOF-MS to distinguish between groups based on the urinary VOC. All of the technologies were consistently able to distinguish between CRC and healthy control with an area under the curve (AUC) greater than 86%). FAIMS was also able to differentiate between early-stage PDAC from healthy and from late-stage PDAC (AUC 89% and 92%, respectively). However, urinary VOC analysis couldn't differentiate PDAC from CP (AUC 0.58). Further studies investigating urinary and faecal VOC of patient for detecting CRC using FAIMS and GC-IMS included 728 subjects which consisted of 26 patients with CRC, 340 with adenoma, 32 polypectomy patients, 296 healthy control and 33 with other gastrointestinal diseases. Both FAIMS and GC-IMS methods were able to separate between CRC and healthy control with AUC above 82%. Patients with adenoma could only be separated from healthy control in faecal odour analysis. Interestingly, odour analysis showed that the faecal VOC profile of patients whom underwent polypectomy returned to its physiological state after three months of polypectomy. This demonstrated the potential of faecal VOC analysis for disease monitoring, especially in adenoma and CRC. Final studies investigated the urinary, breath, and faecal VOC of patients with IBD. Six hundred forty-one subjects in total were recruited: 350 IBD, 276 healthy control, and 15 irritable bowel syndromes (IBS)/Functional Abdominal Pain-Not Otherwise Specified (FAP-NOS). All three types of human waste were consistently able to detect IBD, both Crohn's Disease (CD) and Ulcerative Colitis (UC), from healthy control. Faecal odour analysis was also able to separate between IBD patients with IBS/FAP-NOS, which share many common symptoms. In this work, the potential of utilising urine, breath, and faecal odour-based methods was successfully demonstrated for the non-invasive diagnosis of PDAC, CRC and IBD. The potential use of odour analysis in healthy monitoring was also shown in PDAC patients and patients undertaking polypectomy. Further work in this area will need to focus on sampling and storing samples as well as machine learning approach in order to speed up the advances in the application of modern E-nose technology in a clinical setting. Furthermore, a large-scale multi-centre study involving patients with similar symptoms will push further the knowledge of human waste odour for diagnosis.

This thesis considers qualitative spatio-temporal reasoning (QSTR), a branch of artificial intelligence that is concerned with qualitative spatial and temporal relations between entities. Despite QSTR being an active area of research for many years, there has been comparatively little work looking at large scale qualitative spatio-temporal reasoning - reasoning using hundreds of thousands or millions of relations. The big data phenomenon of recent years means there is now a requirement for QSTR implementations that will scale effectively and reason using large scale datasets. However, existing reasoners are limited in their scalability, what is needed are new approaches to QSTR. This thesis considers whether parallel distributed programming techniques can be used to address the challenges of large scale QSTR. Specifically, this thesis presents the first in-depth investigation of adapting QSTR techniques to work in a distributed environment. This has resulted in a large scale qualitative spatial reasoner, ParQR, which has been evaluated by comparing it with existing reasoners and alternative approaches to large scale QSTR. ParQR has been shown to outperform existing solutions, reasoning using far larger datasets than previously possible. The thesis then considers a specific application of large scale QSTR, querying knowledge graphs. This has two parts to it. First, integrating large scale complex spatial datasets to generate an enhanced knowledge graph that can support qualitative spatial reasoning, and secondly, adapting parallel, distributed QSTR techniques to implement a query answering system for spatial knowledge graphs. The query engine that has been developed is able to provide solutions to a variety of spatial queries. It has been evaluated and shown to provide more comprehensive query results in comparison to using quantitative only techniques.

Background: Klotho (KL) and its paralog β-klotho (KLB) are transmembrane proteins that function as co-factors to aid binding of endocrine fibroblast growth factors (eFGFs) to their respective fibroblast growth factor receptors (FGFR). Klotho has been found to have protective effects at cellular and organismal levels, although the signalling pathways governing these are yet to be determined. Methods: C. elegans strains with genetic deletions in klo-1/KL and klo-2/KL were subject to health and stress resistance analyses. To determine signalling pathways governed by klotho, C. elegans strains with null klo-1/KL and klo-2/KL alleles were crossed into mutants for intracellular signalling components then subject to health and stress analyses. Due to reported impacts of Klotho function in oxidative stress responses, staining for reactive oxygen species was performed in wild-type and klo-2 (ok1862); klo-1 (ok2925) double mutant C. elegans to determine ROS levels in these strains prior to and following stress exposure. Fluorescent reporters for klo-1 and klo-2 gene expression were analysed in wild-type and klo-2; klo-1 double mutant genetic backgrounds before and after pharmacological intervention using reported AMPK activators and TOR pathway inhibitors to elucidate whether AMPK or TOR signalling pathways are impacted by klo-1 and klo-2 deletion. Results: C. elegans with deletion of klo-1/KL, klo-2/KL, or both, have increased survival upon exposure to acute oxidative stress, and show reduced levels of reactive oxygen species (ROS) compared to wild-type counterparts in untreated (control) conditions. Interestingly, in heat stress responses while there is some survival advantage displayed by klo-1/KL or klo-2/KL single mutants compared to wild-type counterparts at the 8 hour time point, deletion of both genes does not increase survival in C. elegans. The survival advantages demonstrated by klotho mutant nematodes is dependent on functional AAK-2/AMPK. Conclusions: In contrast to literature, which suggests removal of klotho would diminish stress responses, C. elegans strains with klo-1/KL or klo-2/KL deletion backgrounds have a survival advantage upon acute stress exposure. This survival advantage is characterised by lowered ROS levels in these animals. The current working theory is that klo-2; klo-1 double mutant mutants could have increased AAK-2/AMPK activity compared to wild-type animals however further research is needed to support this hypothesis.

A measure of tail risk in credit markets is essential to understand the behaviour of credit default swap prices. This thesis presents three tail-risk measures based on dynamic power-law models with multiple time-varying tail parameters. The models use univariate and crosssectional returns of sovereign and corporate credit default swaps to estimate the tail risk at each point of time. The power-law is considered a plausible statistical hypothesis for the tail distribution of returns and measure of tail risk in credit markets. The dynamic power-law exponent is time-varying and persistent. The tail exponent series for 35 European sovereign credit default swaps vary around the mean of 3.0, consistent with the inverse cubic law in other asset classes. Tests show that past exposure to extreme event risk significantly impacts future credit default swap prices and returns. A one-standard-deviation increase in tail risk forecasts an average increase in sovereign credit default swap spreads of 7.6 bps in US credit markets, which is highly significant. These results are robust out-of-sample. The forecasting power of tail risk is also robust to controlling for 25 alternative predictors. Furthermore, tail risk has substantial explanatory power for the cross-section of expected returns in US corporate credit default swaps. Cross-sectionally, firms with high positive loadings on past tail risk earn average expected annual returns 8.1% higher than credit default swaps with low tail risk covariation. Protection sellers increase spreads for credit default swaps bearing high sensitivity to tail risk. This tail risk premium is different from the premiums on market risk, idiosyncratic volatility and coskewness, and robust to considering alternative risk factors. These findings are consistent with asset pricing models that relate tail risk to expected returns and risk premiums.

This thesis includes three papers that helps in identifying, predicting, and reducing the opportunities to commit fraud. Specifically, my first paper focuses on identifying fraud in academic publications; second paper focuses on identifying and predicting fraud using financial statements and third paper focuses on reducing the opportunities to commit fraud by identifying partner characteristics that could improve audit quality. In my first paper I investigate whether Benford's Law can be utilised to determine the validity of data employed within published academic studies. I use the case of Professor James Hunton who had 37 of his articles retracted because there were grave concerns that they contained misstated or fabricated datasets. Based on first significant digits contained in the articles I construct several Benford conformity measures to determine whether Hun ton's retracted papers differ significantly from a control group of non-retracted articles by competing authors. My results clearly indicate that Hunton's retracted papers significantly deviate from Benford Law, relative to the control group of papers. Moreover, my results also suggest Hunton's remaining 19 nonretracted articles may need further examination. My findings suggest both co-authors and journal editors should consider implementing a validation and review process that employs Benford Law as a 'doping' or 'signalling' mechanism with a view to decreasing the risk of fraudulent activity and thereby enhancing the credibility of academic papers and journals. In my second paper, I provide two different models that could help predict financial misconducts. Misreporting continues to have huge impacts on both the domestic and global market. Based on criminology literature the path to financial misreporting is a slippery slope rather than a cliff edge phenomenon and deviations from Benford Law. The first model is based on the proportional hazard model and the second model plots the cumulative deviations and comparing the same with an empirical distribution. Based on prior literature I created a sample of firms that has either received an Accounting and Auditing Enforcement Releases or receipt of a Security 8 Class Action or restatement of financials. I find under both model specifications I can correctly predict the firms that are more likely to misreport in future with a low rate of false positives. In my third paper, I examine the association between the gender of the first-born audit partner's child and audit quality. Based on the protectionism and socialisation theory parental values and preferences are shaped by having a first-born daughter. Specifically, prior literature parents with daughters are found to be more cautious, conservative, risk adverse and adopt many female preferences and values. I investigate and find conclusive evidence that these daughter effects potentially promote characteristics that leads to a higher audit quality. The findings are confirmed by an array of robustness check and thus contribute to the literature that an audit partner's family environment shapes audit outcomes. These findings potentially have direct implications for internal audit firm management as well as investors and regulators.

Livestock have been a cornerstone to human sustenance, wealth, culture, and production for over 10,000 years. Each contemporary livestock species is comprised of a vast assemblage of locally adapted breeds containing unique genetic variation. While most of this variation remains uncharacterised, understanding the genetic diversity is highly valuable in conserving and improving extant livestock breeds. Chapters two to four utilise SNP array data comprised of tens of thousands of loci distributed relatively uniformly across the genome. Firstly, the demographic history of taurine (Bos taurus) and indicine (Bos indicus) cattle was modelled with approximate Bayesian computation (ABC) to determine if domestication occurred across two or three independent events, with model rejection sampling indicating only two domestications took place. Additionally, characterising population structure, admixture, and gene flow to identify migratory events and potential sources adaptive introgression. Secondly, the 15th century colonisation of the Americas by taurine Creole cattle from their putative Iberian ancestors was modelled using ABC. Identifying a founding effective population size (Ne) of 84, followed by a large demographic expansion and subsequent contraction, with higher resolution of recent demographic fluctuations visualised with Ne Slope (NeS) analysis. Furthermore, signals of selective sweeps were scanned for, identifying loci important for tropical adaptation including suggesting a new candidate gene (GDNF) for the slick hair coat phenotype. Thirdly, population structure and selective signals were identified in Ryeland sheep (Ovis aries), which is a resilient and ancient British breed. Haplotype-based detection of selective sweeps and landscape genomics which incorporated environmental, land use, and topographical data identified prominent pathways involved in prion disease pathways and cancer regulation (specifically hepatocellular carcinoma); environmental associations indicate selection may be driven by liver fluke abundance. The final data chapter utilises whole-genome resequencing (WGRS) data to identify SNPs in the feral Chillingham cattle (Bos taurus) herd which has been isolated from gene flow for at least 300 years with minimal management. WGRS bypasses most ascertainment biases inherent to SNP arrays, allowing a more comprehensive characterisation of variation - identifying consistently long runs of homozygosity interspersed with islands of excess heterozygosity. Heterozygous peak windows were putatively maintained by balancing selection and were enriched with quantitative trait loci associated with fertility and milk. Unexpectedly, the major histocompatibility regions were almost absent of SNP variation in Chillingham.

Among species farmed for consumption and trade, parasite epidemics are common due to the intense, often stressful conditions required to maximise production. For aquatic culture, parasite infections are often uncontrollable due to a lack of information and research on both parasites and hosts. Fish lice, ectoparasitic crustaceans, cause severe economic loss to industry each year through physical damage to hosts and subsequent secondary infections. Marine sea lice are notorious in salmon farming and have been the subject of intense research over past decades. Conversely, freshwater lice (Genus Argulus) have been relatively ignored. Here, the problems arising from Argulus spp. infections were investigated with the aim of reducing the economic and health impacts of these parasites. A survey of trout fisheries to ascertain the current state of UK Argulus spp. populations highlighted numerous variables associated with problematic infections and exposed a deficit in effective Argulus spp. control measures. Infection dynamics were then assessed with temperature, host species and infection density all playing a key role in Argulus spp. life histories and generation time. New and novel monitoring and control techniques were evaluated to help tackle infections. Light-baited traps and timed chemical dosage based on circadian rhythms have potential, while egg-laying substrate traps removed millions of Argulus spp. eggs from fisheries to reduce the next generation population size. Finally, a new and invading non-native Argulus species was discovered in UK fisheries and described here morphologically and genetically to facilitate future monitoring of this parasite and improve species identification. Argulus spp. will continue to negatively impact fisheries and aquaculture into the future, however the data and information provided here will help develop integrated management systems to improve fish health worldwide.

In 2020, an estimated 19.3 million new cancer cases were diagnosed, 7.3 % of which were cancer of the prostate. It is evident that the development of prostate tumours is highly dependent on the acquisition of genomic alterations and recent work has identified the potential to stratify patients for therapies based on their mutational status. The phosphatidylinositol-3-kinase (PI3K) signalling cascade is frequently activated in prostate cancer, promoting tumour growth and survival. Additionally, the PI3K pathway has been shown to mediate resistance to androgen-deprivation therapy (ADT), leading to lethal and currently untreatable castrate-resistant prostate cancer (CRPC), thus a key clinical challenge faced in the clinic is how to effectively target the PI3K pathway. Previous work in the lab has identified that Pten loss and Pik3ca activation, that frequently drive PI3K signalling in patients with prostate cancer, cause distinct prostate cancer pathologies and can cooperate to accelerate tumorigenesis and CRPC transition in vivo, suggesting that different genetic drivers within the PI3K pathway can mediate distinct signalling events. To delineate the mechanisms in which oncogenic Pik3ca (Pik3ca+/H1047R) and Pten loss (Ptenfl/fl) in the prostate drive tumorigenic processes, this project utilised RNA-Sequencing (RNA-Seq) and ensemble gene set enrichment pathway analysis (eGSEA). In the intact (uncastrated) setting, we identified significant upregulation of NF-κB transcription factors, pathway mediators and transcriptional targets in Pten-deficient prostate cancer relative to Pik3ca-activated prostate tumours. Our bioinformatic based findings were further validated in Pten-deficient in vitro and ex vivo models of prostate cancer, highlighting the NF-κB pathway as a PI3K-independent avenue for therapeutic intervention in patients with PTEN-null prostate cancer. Furthermore, to determine whether Pten loss in fibroblasts could additionally drive tumorigenic processes in the prostate, this project examined the phenotype of Col1a2CreER(T2) wild-type, Pten+/fl and Ptenfl/fl transgenic mouse prostates. We observed focal prostate hyperplasia within all prostate lobes following heterozygous and homozygous loss of fibroblast cell PTEN, indicating that PTEN loss in either epithelial or stromal cells of the prostate can disrupt normal tissue homeostasis. These findings highlight the need to further understand how PI3K/PTEN signalling coordinates tumour�stroma interactions during prostate tumorigenesis, which may uncover new therapeutic targets and/or predictive biomarkers.

Far from being specialised occurrences, symbiotic relationships including mutualisms, are ubiquitous. The coral reef ecosystem is the product of a mutualistic symbiosis between corals and their endosymbiotic zooxanthellae. The coral polyp creates a sheltered space for the zooxanthellae, while the zooxanthellae provide nutrients to the host through photosynthetic activity. This mutualism has allowed a diverse range of coral animals to evolve and together multiple species of corals create a highly complex 3-dimensional habitat. The plethora of available habitat space has facilitated the evolution of multiple species and, as such, coral reefs are one of the most biodiverse habitats on Earth. In addition, the diversity on coral reefs has facilitated the evolution of a multitude of behavioural niches. Cleaner fish mutualisms are perhaps one of the best-studied interactions on coral reefs. However, the role that the coral habitat plays in facilitating the cleaner-client mutualism is often overlooked and how specific habitat traits affect cleaning behaviour is unknown. This thesis firstly, quantifies the structural components of the cleaner fish's (Elacatinus evelynae) coral habitat and demonstrates that cleaners more frequently inhabit tall corals and that structural complexity increases cleaning interactions (Chapter 2). Unfortunately, due to anthropogenic global change, coral habitat is threatened and corals are disappearing. It is therefore timely that we discover what important attributes corals may contribute to marine inhabitants. Further to this, since the extent of the loss is so vast, multiple studies have sought to artificially replicate coral structure. Although artificial reefs are being widely deployed it is not fully understood how fish respond to these artificial structures at an individual level. Thus, this thesis investigates the behavioural interactions of fish species with novel artificial structures, finding that damselfish associate with artificial structures more than natural corals, potentially due to the increased space available for algae farming (Chapter 3). In addition to the structural element, corals harbour microbial communities that are also threatened by global change. Commensal microbial communities are integral to host health, yet there is little known about these communities in wildlife populations. Fish harbour a diverse microbiota on their skin, and this mucosal layer is in constant contact with organisms in the marine environment; as such it is possible that microbes may be shared between closely interacting individuals. This is particularly relevant to the sharknose goby (E. evelynae), which spends the majority of its time in direct contact with the coral cleaner station. Therefore, the microbial communities of the sharknose goby and its habitat were characterized and intriguingly, cleaners share bacterial genera with Palythoa caribaeorum, a common but toxicbenthic constituent of cleaner fish stations (Chapter 4). Finally, to further knowledge of microbial communities of the sharknose goby, the gut and skin communities are described and using microbial gene predictions several genes associated with toxicity in other fish species are identified (Chapter 5). In summary, this thesis investigates a range of species interactions at multiple scales, from fish and coral habitat interactions to the association of fish with coral bacteria. This thesis furthers our knowledge of how individual fish respond to artificial reefs and demonstrates key aspects of E. evelynae ecology.

In recent years, Structural Health Monitoring (SHM) has attracted significant attention due to its potential in providing effective maintenance strategy for infrastructure. However, interpreting the SHM data remains a challenging task. Model-based interpretation which utilises a behaviour model in the interpretation, requires experts in both developing the model and understanding the change in the model. On the other hand, data-based/model-free interpretation methods reduce the complexity since no physical model is utilised. However, expertise is required in performing data-based interpretation methods due to the need of feature extraction. This thesis is motivated to develop a deep learning-based data interpretation method that can learn features automatically, thereby minimising the required expertise. In this thesis, a deep learning-based method for estimating load capacity of bridges from bridges' images is developed. Data labelling is performed using information from National Bridge Inventory (NBI) database. Parametric study is performed to further investigate the method. A deep learning-based method that utilises correlation between two or more sensor measurements is proposed. This method employs raw measurement data from sensors. The proposed method is implemented for estimating structural responses by using measurements from other sensor as the input. The proposed method is compared with other machine learning methods and the method outperforms the other methods. Two damage detection approaches utilising deep learning techniques are discussed: novelty detection and multiclass classification. Both frameworks successfully predict the presence of the damage that could not be detected by a frequency-based method. An approach that combines deep learning with Moving Principal Component Analysis (MPCA) as an existing damage detection method is introduced. Experimental data collected from a laboratory-scale bridge are employed as a case study to validate the method. A series of investigation on parameters used in both MPCA and deep learning architecture are conducted in order to observe the method.

The advance of Machine Learning (ML) techniques has become the driving force in the development of Artificial Intelligence (AI) applications. However, it also brings about new problems as ML models and algorithms are more data-hungry than ever. First, the sources of data are often geographically distributed and thus gathering all the data for centralised training incurs heavy traffic and prohibitive communication costs. Second and more importantly, moving data out of local devices are now prohibited in many situations due to privacy concerns. As a promising solution, Federated Learning (FL) is a framework proposed by Google for learning from decentralised data without data sharing. Nonetheless, many challenges remain especially regarding how to facilitate efficient FL in different circumstances. The work presented in this thesis is mainly focused on optimising the efficiency of Horizontal Federated Learning, the most commonly used paradigm of FL, from three perspectives: i) redesigning the device- server synchronisation mechanism of FL for failure-tolerant collaborative training over unreliable, heterogeneous end devices (clients), ii) integrating FL with the Mobile Edge Computing (MEC) architecture to utilise the edge layer for robustness, resource efficiency and fast convergence, and iii) optimising the client selection policy in FL through the evaluation of clients' training value based on the learnt representations of their associated data. Results of extensive experiments are presented to demonstrate the effectiveness of the proposed approaches. In addition to algorithm design and experimental observations, this thesis also discusses the research spectrum of FL and overviews its development in the industry including popular platforms, systems and emerging applications. Based on the work presented, key conclusions and insights are provided with a research outlook for the future study of FL.

Cholangiocarcinoma (CCA) is cancer of the bile duct that only has a few treatment options, typically with poor prognostic outcomes. The Proline-Rich Homeodomain protein (PRH), also known as haematopoietically expressed homeobox (Hhex), is a transcription factor that regulates liver and bile duct development. PRH is not expressed in normal bile duct cells but this protein is highly expressed in CCA cells. Using a combination of overexpression and knockdown experiments, PRH was shown to increase cell proliferation, migration and invasion in tumour cell lines. Similar results were obtained in immortalized AKN-1 cells and also in primary biliary epithelial cells. In these cells PRH overexpression also led to the initiation of epithelial to mesenchymal transition (EMT) and anoikis resistance as well as increased cell proliferation and invasion. Gene Ontology and Gene Set Enrichment Analysis of RNA-sequencing data from cells with altered PRH levels reveal that in all cell types PRH regulates multiple signaling pathways and that key genes involved in the Notch signaling pathway, the cell cycle and EMT are regulated by PRH. CCA cells with elevated PRH expression were found to be more sensitive to the CDK4/6 inhibitor Palbociclib as PRH controls the cell cycle through regulation of multiple cell cycle genes including CCND2. Drug screening with a drug repurposing library was performed to identify novel drug treatments for CCA and several repurposed drugs that show a stronger cytotoxic effect when combined with Palbociclib were identified using cell viability assays. This thesis therefore reveals that PRH protein is an important oncoprotein in CCA and that this protein may play a role in initiating CCA and in driving cancer progression and identifies new drug treatment options.

UbcH10 is an E2-conjugating enzyme specific for the Anaphase-Promoting Complex/Cyclosome (APC/C) and a proto-oncogene product that is overexpressed in many human cancers. The molecular basis of UbcH10 function is not, however, completely understood. To investigate the molecular function of UbcH10 in more detail and identify novel UbcH10-interacting proteins, our laboratory had previously performed UbcH10 interactomic studies. As such, it was hypothesised that novel UbcH10-associated proteins might be novel APC/C substrates or APC/C regulators. To this end, we investigated the function of the novel UbcH10-interacting protein and RhoA-specific RhoGEF, PDZ-RhoGEF in more detail. It was determined that PDZ-RhoGEF associates with UbcH10 both in vitro and in vivo and is targeted for proteasomal degradation in early mitosis. In this regard, it was determined that PDZ-RhoGEF associates with Cdc20, at least in part, through a highly conserved KEN box present in PDZ-RhoGEF and that ablation of the KEN box stabilises PDZ-RhoGEF protein levels. Moreover, it was determined that PDZ-RhoGEF is phosphorylated upon a number of S/TP sites in mitosis by Cdk1 and that pharmacological inhibition of Cdk1 in combination with proteasomal inhibitors limits the degradation of PDZ-RhoGEF in mitosis. These data suggest that PDZ-RhoGEF is targeted for APC/C-Cdc20-mediated degradation in a Cdk1- dependent manner. Previous studies have suggested that UbcH10 might possess intrinsic E3 ubiquitin ligase activity. As such it was hypothesized that UbcH10 autoubiquitylation activity might affect UbcH10 function and that UbcH10 E3 ligase activity represented a potential therapeutic target. To this end, it was shown that UbcH10 possesses intrinsic E3 ligase activity and is able to undergo extensive autopolyubiquitylation independent of the APC/C, on one or more lysine acceptor residues. Mutational analyses determined that K119 is the major lysine acceptor residue for polyubiquitin chain formation both in vitro and in vivo. In this regard, it was also determined that UbcH10 can assemble polyubiquitin chains with K6, K11 and K48 homotypic ubiquitin linkages. A high-throughput quantitative UbcH10 ubiquitylation assay for drug-screening was also developed and confirmed the requirement for lysine acceptor residues in UbcH10 autopolyubiquitylation activity. These data indicate that UbcH10 can potentially modify its own activity through self-polyubiquitylation and that UbcH10 E3 ligase activity could potentially be targeted pharmacologically in order to inhibit UbcH10 function.

The existence of bubbles in the nanoscale is an object of debates and disputes in recent years. New types of nanoscale bubble systems have been reported, and among these, Bulk Nanobubbles represent a challenging problem due to their extraordinary stability experimental reported. In pure water, the pressure estimated from the Young-Laplace equation inside a nanobubble of 100 nm diameter will be close to ~30 atm, therefore, questions of their existence/ stability inevitably arise considering that the lifetime of macrobubbles (>1 mm) is on the order of minutes and that of microbubbles (1-1000 µm) is on the order of seconds, whereas bulk nanobubbles (50-300 nm) have been reported to last for days or weeks. Bulk nanobubbles have a higher curvature leading to a higher Laplace pressure, and a larger interfacial area for gas to diffuse out and it should be in principle less stable. Reports stated that they are negatively charged, and speculations arise on the stability mechanism that could affect their stability against the bubble dissolution theories. Bulk nanobubbles are recent, and their research is still in its infancy; however, many applications have been reported and tested, and it appears that there is immense scope for nanobubbles to impact and perhaps revolutionise many current industrial sectors and medical processes.

3D-printing of food is a novel and promising technology capable of creating individualised foods differing in composition, texture, taste or shape. Objects are created layer-by-layer and the most common additive manufacturing technique for food relies on an extrusion process. In order to formulate for structure and texture of food objects via a printing process, in-depth understanding of the solidification mechanism and the material properties is of highest importance. While most of the current research on extrusion printing focuses on pastes and gels, which maintain their shape due to a yield stress, the key challenge of this work lies in a fast, local and irreversible transition of the feedstock from a solution (sol) to a gel. High dairy protein-based materials, which offer health benefits and have high satiating effects, were used to create and tailor novel edible formulations for applications in thermal extrusion printing. First, characterisation of a solidification mechanism suitable for thermal printing of high dairy protein-based feedstock including a fast sol−gel transition was conducted. Investigation of the pH−temperature (T)-route, cold acidification followed by heating, on the rheometer provided evidence that the temperature-time profile used to trigger gelation was comparable to 3D-print-ing but without superimposed flow. The sol-gel transition temperature as determined by rheology was used to mimic dynamic conditions during printing by applying a steep temperature gradient. Firm gels were produced and classified regarding their physical properties and status around the sol-gel transition. Formulations showing preferable characteristics, including a low storage modulus G' (~ 0.1 Pa) at low temperatures (2°C) followed by a steep increase of G' during sol-gel transition, were considered to be suitable feedstock for applications in thermal extrusion 3D-printing. Second, the two main dairy proteins (micellar casein and whey protein) were combined to create and to tailor novel edible casein−whey protein suspensions. These formulations were characterised regarding their potential use for extrusion-based 3D-printing and several design rules ii were established. Sol−characteristics at low temperature (2°C) were confirmed, providing a liquid feedstock which could be transferred into a gel via the pH−T-route. The pH at heating during denaturation of the whey proteins as well as the overall protein content were crucial parameters to change the microstructure of casein−whey protein suspensions, e.g. resulting in unwanted pre−gelation (G' > 1 Pa) at too high protein contents. Formulations with depleted casein micelles (CM), caused by desired dissociation of κ-casein at increased pH at heating, showed lower sol−gel transition temperatures and increased aggregation rates. The aggregation rate (represented by the storage modulus G') was proposed as a first good indicator for the printability of high dairy protein-based formulations. Furthermore, for the creation of more complex and novel edible formulations, dairy fat was added to casein−whey protein suspensions. The additional product parameter, fat, changed the microstructure as the overall total solid content of the formulations increased. Mechanical treatment was necessary to obtain smaller milk fat particles (< 1 μm) covered with protein, which could mimic the protein behaviour and actively contribute as pseudo-proteins to the gelation process via the pH−T-route. With increasing pH at heating, more depleted casein micelles were found on the surface of mechanically treated fat globules, causing a decrease in the sol−gel transition temperature and an increase in the aggregation rate. At a higher pH at heating, more particles contributed to the gelation per unit area and the surface properties of the fat globules covered with certain proteins, e.g. κ-casein depleted CM, were favourable, shown via electrophoresis and a newly developed protocol for low dose transmission electron cryogenic micros-copy. Several protein−fat formulations with promising aggregation rates could be printed. Overall, this thesis demonstrates that high dairy protein-based formulations with and without additional fat can be used as a novel feedstock for thermal extrusion 3D-printing applications via the pH−T-route, if product- and process parameters are precisely adjusted, and all the established design rules are adhered to.

Cancer cells exhibit a rewired metabolism to meet their specific energy needs for cell proliferation. Despite remarkable efforts to unravel how cancer cells reprogram their metabolism, characterization of the metabolic differences between the different subtypes of cancer has not been well performed. Haematological cancers can be divided into subtypes with significant implications for diagnosis and treatment. The metabolic profiles of a panel of 18 haematological cancer cell lines representing acute myeloid leukaemia (AML), chronic myeloid leukaemia (CML), multiple myeloma (MM), Burkitt's lymphoma (BL) and diffuse large B cell lymphoma (DLBCL) were characterised using an NMR approach. I found a notable widespread metabolic heterogeneity between haematological cancer subtypes, even though all haematological cancer cells showed a typical Warburg effect with lactate production. Intriguingly, MM cells were observed with a large amount of leucine uptake. The exploration of leucine metabolism using stable isotope tracer revealed that leucine is fully metabolised in MM cells and feeds the TCA cycle as well as sialic acid formation. Moreover, metabolic differences between BL and DLBCL were scrutinised, as these diseases tend to share similar morphological, immunophenotypic and genetic characteristics. Exogenous asparagine was found to regulate serine metabolism in BL cells. Furthermore, the combination of asparaginase (ASNase), a clinically used agent that depletes exogenous asparagine, with a PHGDH inhibitor (NCT503) had a synergistic effect on the viability in BL cells, exhibiting more decrease in cell viability as compared to those obtained from treatment of ASNase alone. Conversely, the combination of ASNase with NCT503 showed no synergistic effect on the viability of DLBCL cells, suggesting that the combination of ASNase with NCT503 is solely synergistic for BL. Hence, NMR-based metabolite profiling has revealed how metabolism alters in haematological cancers. This thesis has highlighted and investigated metabolic dependencies and vulnerabilities in different haematological cancer which can be utilised for diagnostic and therapeutic purposes in future.

Encapsulated sorbents are emerging as a new tier of composite sorbent materials that furnishes liquid sorbents with the properties of solid sorbents. The favourable properties of selectivity and high loading capacity of liquid sorbents are further enhanced by the discretization into smaller encapsulated units of higher surface area which also reduce solvent loss and retain precipitates and by-products. Recently, encapsulated sorbents have been demonstrated in applications for the removal of CO2 from effluent gases using advanced solvents with the potential of increasing the energy efficiency of the process with respect to established technologies. Sorption thermochemical energy storage systems could equally benefit from encapsulation where liquid sorbents consisting of aqueous solutions of hygroscopic salts are prone to crystallization and corrosion issues. It is immediately clear that for such composite materials, shell material development should be geared towards selectively permeable shell materials able to retain the hygroscopic salts whilst allowing exchange of the solvent (water) with the environment. Additionally, such shell materials must have sufficient mechanical strength to withstand rupture as well as thin walls to minimise mass transfer resistances. Moreover, the impact of the deformability of the encapsulated sorbents under external loads, such as in packed beds, on the sorption performance requires greater investigation.

Traumatic injuries are a leading cause of death worldwide. Whilst the physical injuries can be treated patients are still at risk of developing an infection and becoming septic. It is hard to diagnose sepsis in patients of trauma as the inflammatory response to sepsis is masked by the inflammatory response to the traumatic injury. Current diagnostic techniques involve culturing blood samples or identifying a site of infection, neither of which are fast enough to allow for the best patient outcome. As such there is much need for a point of care diagnostic test. This work attempted to tackle this problem in two ways. First by analysing a metabolomics data set to identify biomarkers of sepsis in burns patients. Second, by developing a quantitative fluorescent assay for the detection of kynurenine, a biomarker of sepsis. The data analysis was performed on a metabolomics data set acquired from clinical samples from burns patients. Three different classifiers were used to create models on the data set, k-nearest neighbours, naive Bayes and logistic regression. Classifier performance was above average on the data set, with the k-NN technique providing an AUC value of 0.82 and sensitivity and specificity of 85 % and 70 % respectively on the early-sample class-balanced subset of the data. The t-test and minimum redundancy maximum relevancy (MRMR) feature selection techniques were performed on the data set along with lasso with logistic regression. The t-test and logistic regression identified glucose and lactate as being the most important features whereas MRMR identified glucose and not lactate. These are both already well known biomarkers used in controlling the outcome of sepsis patients. A literature search identified the metabolite kynurenine as being a positive biomarker for sepsis in patients of trauma. A fluorescent molecule was synthesised which upon binding to kynurenine causes a large bathochromic shift in the emission spectrum of the fluorescent sensor from 470 nm to 560 nm. A fluorescent assay was created using the standard addition technique to quantify the amount of kynurenine in a urine sample. The standard addition technique was chosen as it removes matrix affects which would be present when using biological samples. Tests performed on pure samples of kynurenine, gave results within 5 % of the actual concentration. Synthetic urine was then used, giving results within 10 % of the actual concentration. HPLC experiments were performed to quantify the amount of kynurenine in actual urine samples. The HPLC protocol was validated with kynurenine in solutions of water and synthetic urine, however complications arose when using urine which meant the clinical samples could not be accurately quantified. As such, a qualitative experiment showed linear fluorescence plots when performing the standard addition technique with actual urine samples. The results indicate the assay will work in urine. Being fluorescence based it has the potential to move into a point of care device. Future work would go towards quantifying the kynurenine concentration in clinical samples and then optimising the assay parameters once known concentrations can be determined.

Efficient heat recovery and management is critical for the economic and environmental performance of a micro-CHP system. The cathode air pre-heater (CAPH) plays a pivotal role in the operation of Solid Oxide Fuel Cell micro-CHP systems as it heats up the inlet air close to the operating temperature (e.g. 750°C) of the SOFC by utilising the recovered thermal energy from the high temperature exhaust gas of the SOFC system. The current CAPH stacks, which are made from high Cr-containing alloys, are still vulnerable to high temperature corrosion and chromium evaporation, which can have a serious detrimental effect on the lifetime of the SOFC stack. For this reason, the aim of this study was to perform extensive corrosion and chromium evaporation measurements on various high temperature alloys and to implement a material progression for CAPH from chromia-forming alloys to alumina-forming alloys. The effect of exposure temperature and water content on the oxidation and Cr evaporation behaviour of Inconel 625, SS309 and AluChrom 318 was analysed. The findings of this study showed that Cr evaporation and oxidation rates of Inconel 625 and SS309 were dramatically reduced when the temperature was reduced from 850 to 650°C. AluChrom 318 also exhibited a decreased oxidation rate with decreasing temperature but it also demonstrated a reverse trend in Cr evaporation temperature dependence. The main effect of water vapour on the alloys exposed to high temperature was to accelerate Cr evaporation. The corrosion behaviour and Cr retention capability of an aluminised alloy (aluminised SS309) and an alumina-forming alloy (AluChrom 318) were compared with a conventional chromia-forming alloy (SS309) in simulated SOFC cathode atmosphere. The SS309 samples failed the specifications for application as cathode air preheater (CAPH) material due to a high Cr evaporation rate and massive scale spallation. The surface aluminisation dramatically decreased the Cr evaporation rate of SS309, but not in the long-term due to breakaway oxidation induced alumina scale spallation. The AluChrom 318 formed a γ-Al₂O₃/α-Al₂O₃ layer scale and demonstrated the highest Cr retention ability and oxidation resistance among the tested alloy samples. Microstructural analysis was conducted on an AluChrom 318 CAPH stack after 25,000 hours of industrial operation in simulated SOFC CAPH environment. The results showed a large amount of Cr₂O₃ formation around the exhaust outlet (cold zone) and a high Al oxidation rate around the exhaust inlet (hot zone). This implied that the as-received AluChrom 318 CAPH plates did not have sufficient corrosion protection and a good barrier layer to stop Cr evaporation formed at high temperatures. For material optimisation, pre-treatment of AluChrom 318 in a high temperature oxidising atmosphere was then prescribed to achieve a better oxidation resistance and Cr retention capability of CAPH stacks in service. It was found that pre-treatment applied on the AluChrom 318 developed an α-Al₂O₃ surface scale that reduced the Al oxidation rate and reduced the tendency towards formation of a Cr₂O₃ layer.

When Zirconium claddings are exposed to irradiation and oxidation in reactor, after a certain period, they may undergo a change of hydrogen uptake behavior and in some cases an increased hydrogen pickup fraction. Hydrogen is present in reactor as a by-product of the oxidation, or through radiolysis of the coolant. It may penetrate through the oxide and enter the underlying metal. Above a certain concentration, it precipitates as brittle Zr-hydrides degrading the cladding's mechanical properties. The work performed in this study contributes to the improved understanding of the hydrogen uptake mechanism in reactor. Parameters, which have been suggested to contribute to the changes in hydrogen uptake at high burnups, have been clearly quantified.

Emerging evidence suggests that glial cells that have a role in modulating inflammation, play an active role in the pathogenesis of Neurodevelopmental and Psychiatric Disorders (NPDs), including autism spectrum disorders (ASD) and schizophrenia (SCZ). Genome-wide association studies have shown that chromosomal mutations such as Copy Number Variation (CNV) at 1q21.1 loci (deletion and duplication) are associated with significant risk for NPDs. However, the impact of this mutation on glial development and physiology and how it contributes to increased risk for NPDs remains unknown. Using patient-derived iPSCs carrying CNVs the role of 1q21.1 chromosomal locus on the functionality of neuroimmune cells-astrocytic and microglial and its contribution to disease pathology has been explored in this thesis. The result presented in this thesis demonstrated that 1q21.1 CNV is associated with altered astrocyte morphology and its effect on functional abilities like phagocytosis, glutamate uptake, ATP production, and calcium activity. Further, RNA-seq data from astrocytes carrying 1q21.1 CNV had highly significant changes in genes linked to ASD and SCZ. To study the impact of 1q21.1 CNV carrying astrocytes on neuronal health, a co-culture system was developed where astrocytes carrying 1q21.1 CNV were cultured with healthy neurons. The co-culture of healthy neurons with 1q21.1 CNV carrying astrocytes resulted in differential expression of synapse associated genes suggesting an impact on synapse formation in healthy neurons. Further, the co-culture of astrocytes carrying 1q21.1 CNV with healthy neurons resulted in increased production of reactive oxygen species, suggesting compromised neuroprotective function of astrocytes. The network activity of these co-cultures also demonstrated a deficit compared to healthy neuronal network activity as quantified by network firing rate and synchronization. To establish the role of critical genes in the 1q21.1 CNV, two knockout lines (KO-GJA5 and KO-GJA8) were generated using CRISPR technology. The functional assays performed on these lines mimic similar results as in astrocytes carrying 1q21.1 deletion. To further investigate the 1q21.1 CNV on neuroinflammatory status, the microglia like cells were derived from iPSCs. The pathway analysis of the microglia carrying 1q21.1 CNV demonstrated differential expression of essential immune and inflammation regulatory genes. Also, time and dose-dependent cytokine production were compared with human microglia. Further, the iPSC derived microglia were also used to understand the immunomodulation following traumatic brain injury. The result presented in this thesis for the first time demonstrates the contribution of 1q21.1 loci to NPDs by glial dysfunction and immunomodulatory pathways.

Many woodland bird species within Britain have shown population declines over recent years, with unclear reasons for declines. The Hawfinch (Coccothraustes coccothraustes) has been declining within the UK since the 1970's, while mainland Europe populations have remained stable. This PhD thesis used DNA metabarcoding and high-throughput sequencing (HTS) of Hawfinch faecal samples to describe plant and invertebrate dietary composition across core Hawfinch population ranges within the UK and mainland Europe. I investigated the degree of dietary composition difference between distinct populations of Hawfinch as well as between demographic groups. This was to elucidate the extent that Hawfinch show dietary plasticity, which can be a determining factor allowing species to adapt to environmental changes. Given the importance of diet, it is important to understand which taxa are preferred, especially for species which are showing population declines. Hawfinch plant dietary composition was analysed to reveal if the frequency of plants detected within their diet differed from their foraging environment, indicating selective foraging. UK Hawfinch dietary composition of plant (Chapter 2) and invertebrate (Chapter 3) taxa was found to vary spatially and between demographic groups, with mainland Europe populations showing similar patterns (Chapter 4). Analysis of the relative abundance of herbivorous taxa within UK woodlands compared with frequency of detection within Hawfinch diet (Chapter 5) indicated Hawfinch were showing selective foraging and were not consuming certain taxa relative to their availability. This PhD thesis provides novel insights into the dietary plasticity of a declining species across a broad geographical range. I provide a clear example of how DNA metabarcoding methodologies can be applied in studies of difficult to study species, as well as how dietary composition can be driven by environmental and demographic factors. Finally, this thesis shows that Hawfinch are selective, and are consuming plants disproportionally to their availability.

Lyme disease is the most common vector-borne infection across temperate zones of the Northern Hemisphere and is caused by the spirochete, Borrelia burgdorferi sensu lato (s.l). The outer membrane (OM) of Borrelia possesses many unique features, including an abundance of lipoproteins and few integral β-barrel membrane proteins. The predicted small β-barrel OMPs from Borrelia burgdorferi s.s; BB_0027, BB_0405, BB_0406 and BB_0562, are all predicted to form an 8-stranded transmembrane β-barrel. Three truncations were made to each OMP to increase the chance of crystal formation and these, along with potentially soluble targets, were cloned and recombinantly expressed in E. coli. Following purification, a BB_0406 truncation was soluble in DDM and showed promise for crystal formation. Factor-H binding was investigated using a far-Western blot. No binding was detected, in agreement with the most recent literature. The Borrelia genome consists of a linear chromosome and numerous linear and circular plasmids. While the chromosome of Borrelia has been subject to various searches for β-barrel OM proteins, the plasmids have so far remained unexplored. This research utilised a computational-framework approach to identify potential OM β-barrel proteins in the plasmid proteome of B. burgdorferi s.l. This approach identified two plasmid-encoded proteins: the annotated porin OMS28, and a previously uncharacterized protein BBJ25. Orthologs of BBJ25 are detected in two other major spirochaete families, Brachyspira and Treponema. Within the Borrelia species complex, BBJ25 is found within a predicted 7-gene operon at either the 3rd or 7th position, resulting in two allelic variants. Phylogenetic analysis of the allelic variants is inconsistent with vertical descent alongside the chromosome suggesting that horizontal transfer and/or recombination has occurred after the divergence of Relapsing Fever/Lyme Disease-type Borrelia. While the precise role of BBJ25 remains unknown, the identification of a novel conserved outer membrane protein is of interest as a diagnostic or therapeutic target.

The manufacturing industry is undergoing significant transformation towards electrification (e-mobility). This transformation has intensified critical development of new lightweight materials, structures and assembly processes supporting high volume and high variety production of Battery Electric Vehicles (BEVs). As new materials and processes get developed it is crucial to address quality defects detection, prediction, and prevention especially given that e-mobility products interlink quality and safety, for example, assembly of 'live' battery systems. These requirements necessitate the development of methodologies that ensure quality requirements of products are satisfied from Job 1. This means ensuring high right-first-time ratio during process design by reducing manual and ineffective trial-and-error process adjustments; and, then continuing this by maintaining near zero-defect manufacturing during production by reducing Mean-Time-to-Detection and Mean-Time-to-Resolution for critical quality defects. Current technologies for isolating and mitigating quality issues provide limited performance within complex manufacturing systems due to (i) limited modelling abilities and lack capabilities to leverage point cloud quality monitoring data provided by recent measurement technologies such as 3D scanners to isolate defects; (ii) extensive dependence on manual expertise to mitigate the isolated defects; and, (iii) lack of integration between data-driven and physics-based models resulting in limited industrial applicability, scalability and interpretability capabilities, hence constitute a significant barrier towards ensuring quality requirements throughout the product lifecycle. The study develops a transformative framework that goes beyond improving the accuracy and performance of current approaches and overcomes fundamental barriers for isolation and mitigation of product shape error quality defects in multi-station assembly systems (MAS). The proposed framework is based on three methodologies which explore MAS: (i) response to quality defects by isolating process parameters (root causes (RCs)) causing unaccepted shape error defects; (ii) correction of the isolated RCs by determining corrective actions (CA) policy to mitigate unaccepted shape error defects; and, (iii) training, scalability and interpretability of (i) and (ii) by establishing closed-loop in-process (CLIP) capability that integrates in-line point cloud data, deep learning approaches of (i) and (ii) and physics-based models to provide comprehensive data-driven defect identification and RC isolation (causality analysis). The developed methodologies include: (i) Object Shape Error Response (OSER) to isolate RCs within single- and multi-station assembly systems (OSER-MAS) by developing Bayesian 3D-convolutional neural network architectures that process point cloud data and are trained using physics-based models and have capabilities to relate complex product shape error patterns to RCs. It quantifies uncertainties and is applicable during the design phase when no quality monitoring data is available. (ii) Object Shape Error Correction (OSEC) to generate CAs that mitigate RCs and simultaneously account for cost and quality key performance indicators (KPIs), MAS reconfigurability, and stochasticity by developing a deep reinforcement learning framework that estimates effective and feasible CAs without manual expertise. (iii) Closed-Loop In-Process (CLIP) to enable industrial adoption of approaches (i) & (ii) by firstly enhancing the scalability by using (a) closed-loop training, and (b) continual/transfer learning. This is important as training deep learning models for a MAS is time-intensive and requires large amounts of labelled data; secondly providing interpretability and transparency for the estimated RCs that drive costly CAs using (c) 3D gradient-based class activation maps. The methods are implemented as independent kernels and then integrated within a transformative framework which is further verified, validated, and benchmarked using industrial-scale automotive sheet metal assembly case studies such as car door and cross-member. They demonstrate 29% better performance for RC isolation and 40% greater effectiveness for CAs than current statistical and engineering-based approaches.

Concussion occurs when a "biomechanical force, transiently disturbs normal brain function causing neurological, cognitive and behavioural signs and symptoms". However, the diagnosis of concussion is imprecise and challenging. Currently, there is a great deal of attention on concussion in contact sports due to the immediate and potential long-term effects on the brain. The transient and complex array of symptoms that occur, mean patient compliance is essential, to quantify the neurological deficit that enables an experienced assessor to diagnose accurately. The future in concussion diagnosis therefore lies in the development of a simple, relevant, rapid and objective assessment of the trauma to the brain. The most likely candidates are either brain imaging e.g. MRI, or measurement of a relevant biomarker or group of biomarkers. Rapid diagnosis is particularly needed in contact sports such as rugby and football to enable players to safely return to play. The development of an objective diagnostic test may also help to explain the longer-term impact of head injury, and be applicable to other sports and situations e.g. military or household accidents etc. In this thesis, I have hypothesised that changes in small non-coding RNAs (sncRNAs) and extracellular vesicles (sEVs) in saliva may be used for the diagnosis of concussion. Changes in either or both of these in saliva could reflect rapid changes in the brain and systemic circulation due to the concussion. Understanding of mechanism through which these changes could occur is beyond the scope of this thesis. The Study of Concussion in Rugby Union through MicroRNAs (SCRUM) (ISRCTN16974791) has used next Generation Sequencing (NGS) and qPCR validation steps to analyse 390 saliva samples from the top two tiers of professional rugby in England. Samples were collected during training, 3 hours and 48-72 hours post-injury. Controls were taken from groups with musculoskeletal injury, uninjured players, and players in whom concussion had been suspected but ruled out on. I have identified 12 sncRNAs that undergo up-regulation in saliva of players with concussion (ten in the immediate aftermath, two 48-72 hours post injury). Two sncRNA families had more than one member responding to injury, namely the RNU-6 snRNA and Let-7 miRNA families. Further work is required to confirm whether measurement of these, or a combination of sncRNAs, can be used to diagnose concussion. sEVs were measured in saliva using a new assay, the ExoView, but no clear pattern of change was observed. This is likely to reflect the complexity of sEVs which vary markedly in size and can be generated from multiple cell types. Studies on miRNAs, were extended to a deceased group of patients who had been diagnosed with Alzheimer's disease (AD), in view of a possible historical link to concussion. Brain samples were obtained from the South West Dementia Brain Bank (SWDBB, University of Bristol). Two progressive stages of AD and a control group were assessed using qPCR. The Let-7 family miRNAs, hsa-let-7i-5p and hsa-let-7a-5p, were found to be differentially expressed in the brains of AD patients. Further work is required to understand the function of these two mRNAs in the brain. In conclusion, this study has identified several sncRNAs that may change in saliva in response to concussion and which could form the basis of a future diagnostic test. This work has also provided preliminary evidence of changes in the Let-7 family of miRNAs in AD. Further work is required to understand the significance of changes in sncRNAs in concussion and AD, and a possible link between their role (if any) in these two disorders.

In this thesis, I investigate the contribution of human papillomavirus (HPV) induced epigenetic changes to the process of cervical carcinogenesis. I used an in vitro progression model of cervical cancer involving primary human foreskin keratinocytes in conjunction with human cervical intraepithelial neoplasia cell line and different cervical cancer cell lines. The Primary Human Foreskin Keratinocyte from one donor transfected with episomal HPV16 and HPV18. Following propagation of PHFK HPV18 and PHFK HPV 16 cell lines, I have shown HPV-induced cytopathic changes in human keratinocytes using immunocytochemistry. I found that HPV-associated with changes in the expression of key epigenetic regulators that follow the HPV life cycle and the process of cervical carcinogenesis. Using PCR methods I showed that specific regions of HPV16 and HPV18 genome are prone to disruption during HPV integration. Overexpression of HPV E7 oncoprotein precedes the disruption of full length HPV E2 and induces expression and enzymatic activity of DNMT1 and DNMT3B. Using co- immunoprecipitation, I found that the endogenous HPV E7 oncoprotein interacts at physiological level with DNA methyltransferase 3B. I have also shown, using chromatin immunoprecipitation, that HPV associated with pattern of epigenetic switching from histone 3 lysine 27 trimethylation to DNA methylation. Using a genome wide methylation array, I found that HPV18 integration induces changes to the human DNA methylome similar to the changes expected in cervical neoplasia. Finally, I show that methylated tumour suppressor genes following HPV integration are reversibly demethylated with 5-azacitidine.