Search

This thesis investigates the consumers' motivations to buy, and estimation of the willingness to pay for, organic food and the environmental attributes associated with it. These attributes include environment friendliness, animal welfare, low chemical usage, health, and quality, all shown to be significant parameters of social welfare. The thesis estimates results from revealed, stated and combined data, addressing the issues associated with single estimations, above all hypothetical bias, collinearity and endogeneity of attributes, and data scale differences. The true reasons why consumers appreciate environmental attributes, and their driving behaviour, are imperative from a marketing, industry and policy perspectives.

It is generally accepted that education is a powerful enabler of social mobility in the modern world. However, access to quality and specialized education is still highly unequal, creating pathways for some groups of people and barriers to others. This dissertation explores three approaches to understanding and addressing inequalities of opportunity in three separate chapters. The first chapter examines the importance of test design, in particular the time limit component, as a driver of gender gaps in performance. The second chapter evaluates the potential of a growth mindset intervention for narrowing gender gaps in challenge-seeking and competitive behavior. The third chapter investigates the educational attainment and labor market outcomes of an affirmative action policy for college admission that targets low-income and underrepresented racial minorities. Respective results show that changing the time limit of a test may increase female representation in competitive programs, teaching growth mindset may help some women become more challenge-seeking and competitive, and that affirmative action can more than double the chances of black low-income students entering a prestigious career without negatively impacting the prospects of students displaced by the policy. In summary, these studies provide evidence that inform the potential of an institution, a practice, and a policy to open pathways for more equitable opportunities in education and the labor market. [The dissertation citations contained here are published with the permission of ProQuest LLC. Further reproduction is prohibited without permission. Copies of dissertations may be obtained by Telephone (800) 1-800-521-0600. Web page: http://www.proquest.com/en-US/products/dissertations/individuals.shtml.]

Traditional supervised machine learning techniques need to be adapted when applied to longitudinal datasets, due to their specific characteristics such as a large amount of missing data and the dependency between repeated measurements of the same variables. These adaptations range from data preprocessing techniques that maintain and use information from the underlying temporal data structure of longitudinal datasets, to algorithm adaptations that consider the temporal aspect of the data when making predictions. In this thesis we focus on the classification task of supervised learning, in the context of longitudinal biomedical and health data from ageing studies. More specifically, we address the problem of predicting the diagnosis of age-related diseases, given several years of observations about each instance (individual). In order to evaluate our proposed approaches for longitudinal supervised learning (described below), we created 30 longitudinal classification datasets. These datasets are comprised of data from the English and Irish longitudinal studies of ageing, which collect biomedical and self-reported health information on thousands of participants, over multiple waves carried out throughout the years. Regarding supervised learning algorithms, we focus on decision tree-based algorithms, namely Random Forests (which learn an ensemble of decision trees) and a decision tree algorithm (which learns a single decision tree). These algorithms were chosen because they represent a good trade-off between predictive accuracy and interpretability, which is particularly relevant for our health application. Random Forests are known to achieve high predictive accuracy in general, and are partially interpretable (via feature importance measures), whilst decision trees are directly interpretable, although usually less accurate than Random Forests. This thesis' main contributions are three new approaches for coping with longitudinal data in supervised learning (particularly classification). The first two main contributions involve data preparation, namely missing value replacement and the construction of features representing temporal information in the data. These contributions are independent from the choice of classification algorithm to be applied to the longitudinal data, so they are widely applicable to longitudinal studies. The third main contribution involves algorithm adaptation, adapting decision tree-based algorithms to consider the temporal information in the data. More precisely, the first main contribution of this thesis is the proposal of a data-driven missing value replacement approach to estimate the missing values in longitudinal datasets. The proposed approach performs a feature-wise ranking of an input set of missing value replacement methods, using known data as ground-truth to estimate the error rates of each method. Then it uses that ranking to choose the best missing value replacement method for each feature. Experiments have shown that this approach improved predictive accuracy in general, by comparison with several baseline methods for handling missing values. The second main contribution consists of several types of constructed temporal features, which are are calculated (in a data preprocessing phase) using the repeated measurements of the original longitudinal features. These constructed features represent different types of temporal patterns that can occur in longitudinal datasets. The constructed features are then added to the original dataset, and used together with the original features when running any chosen classification algorithm. Experiments have shown that the constructed features benefited from datasets with more temporal data available, and that the added features overall increased the predictive accuracy of the Random Forest classifiers. The third main contribution of this work is an algorithm adaptation approach for decision tree-based algorithms (more precisely, Random Forests and decision tree algorithms) applied to longitudinal data inputs. We adapted the node split function of such algorithms to consider two criteria, using a lexicographic optimisation approach. This approach first tries to select the best split feature at each tree node based on the features' information gain ratio, as the primary criterion. If, however, two or more features have about the same information gain ratio, as a tie-breaking (secondary) criterion, the algorithm prefers to select a more recent feature, since these are assumed to be more relevant for classification than older features. Experiments have shown that this lexicographic split approach led to increased predictive accuracy in general for the Random Forest classifier.

Deep learning is a particular kind of machine learning which is powerful and flexible as a consequence of its ability to represent the world as a nested hierarchy of concepts (LeCun et al. 2015, Goodfellow et al. 2016). Due to the ever increasing power of parallel computing graphics processing units, larger labelled datasets and improved training techniques, great leaps in the performance of various machine learning tasks have been achieved using deep learning (LeCun et al. 2015). At the time of this writing, deep learning is the dominant machine learning approach for much ongoing work in fields such as: Computer Vision (Krizhevsky et al. 2012, He et al. 2016), Reinforcement Learning (Mnih et al. 2015, Silver et al. 2016), Medical Imaging (Ronneberger et al. 2015), and Natural Language Processing (Vaswani et al. 2017, Devlin et al. 2019). However, with the uptake of deep learning models into safety-critical domains, transparency of model predictions is becoming increasingly important for: safety, decision-making, fairness and legislative reasons. Moreover, designing deep learning models that strike a good balance between human interpretability and performance has proven to be a challenging task (Caruana et al. 2015, Montavon, Lapuschkin, Binder, Samek & Müller 2017, Kendall & Gal 2017, Rudin 2019, Samek et al. 2019). With that said, in this thesis we advocate for an alternative view of interpretability based on estimating the uncertainty in a model's predictions, which serves as a proxy for model transparency. In our investigations, we formalise the desiderata of model transparency as: trust, information and generalisation, and take steps towards the development of deep learning models which have the potential to satisfy them. Concretely, we leverage the language of uncertainty to improve the performance and transparency of deep learning models in computer vision tasks, providing probabilistic techniques to enhance more interpretable models by design such as capsule networks.

Cardiac contractility is often defined as the ability of the cardiac muscle to generate power and results from a number of mechanisms. Contractile work is also the trigger that causes an increase or decrease in ATP generation, via one of two main processes: fatty acid (FA) oxidation or carbon substrate metabolism. These two mechanisms are crucial in contraction and any changes can lead to cardiac dysfunction. To this effect, both age and obesity have been previously shown to play a significant role in cardiac dysfunction, primarily via changes to contractility and ATP generation and in pathways involving the citric acid cycle and uncoupling proteins. The primary aim of this thesis was to investigate cardiac contractility in rodent models of aging and high fat diet-induced obesity (HFD) in the presence and absence of inotropic drugs. The first part of this project was a physiological investigation into the effects of ageing (3, 6 and 18-month murine models) and high fat diet-induced obesity (6 and 18-month lean and obese murine models) on cardiac function. The second part of the project was a pharmacological investigation into the effects of inotrope-induced changes on cardiac contractility in ageing and high-fat diet-induced obesity (using the same models). To achieve this, three techniques were used: the Langendorff isolated heart model, the papillary muscle work-loop assay used to assess power output and western blotting protein analysis of Pyruvate dehydrogenase E1-α subunit (PDH) and Uncoupling protein 3 (UCP3). Cardiac function of the whole heart diminished with age, showing significant differences for the Rare pressure product (RPP) on the 18-month models when compared to the 3 and 6-month models (p<0.05), but the cardiac muscles also suffer time and age-dependant decreases in performance, when comparing the 6-month models to the 3-month models. This translated to reductions in power output (p<0.05) and total net-work done. In addition to this, and in agreement with previous literature, the results from the western blots on PDH have shown decreased phosphorylation in an age-dependant fashion and indicate that there is, in fact, a decrease in FA oxidation (p<0.0001 when comparing between the 6 and 18-month models with the 3-month model). The UCP3 results corroborate these findings and suggest that there is a decrease in the rate of FA oxidation, in an age-dependant fashion (p<0.0001 when comparing between the 6 and 18-month models with the 3-month model). HFD functional changes were also recorded, as the 6-month lean heart showed significant decreases in the absolute values for the haemodynamic parameters when compared to the 6-month HFD model (p<0.0001); the 18-month model showed a significant increase compared to the 6-month lean model, but not when compared to the 6-month HFD model, which aligns with previous literature which associated these changes with hypertrophic changes to the heart. On the muscle assay, significant decreases in power output, fatigue and total net-work done were recorded when comparing between lean and HFD muscles (p<0.01). Finally, the pPDH and UCP3 were shown to be effective biomarkers of obesity, as the formers levels decreased in the presence of the high-fat diet (p<0.0001), while the latter increased significantly. The measured increase in UCP3, specifically, is indicative of increased FA availability and its consequent usage. On the ageing pharmacological chapter, a phenomenon known as β-receptor desensitization was present, as the LVDP, HR and RPP for the 18-month dobutamine-treated hearts were shown to have age-dependant impairments, when compared to the 3 and 6-month models (p<0.0001). As for the western blot assay, it was found that PDH was reduced in an age-dependant fashion for all three drug-treated hearts, with seemingly no changes in expression between them (p<0.0001 when comparing between the 6 and 18-month models with the 3-month model). This seems to indicate that FA oxidation and ATP production are being affected, but not detected in the isolated heart models (other than on the Dobutamine-treated hearts). As for UCP3, interestingly no change was recorded for atenolol or itraconazole across different ages, but a significant decrease was recorded for the 6 and 18-month dobutamine treated hearts when compared to the 3-month heart (p<0.0001). It is therefore possible that there is a correlation between the observed β-receptor desensitization and the UCP3 levels, even though the exact link between the two is unclear. On the isolated heart model, between the lean and HFD atenolol-treated hearts, but both the Dobutamine and Itraconazole data showed severe impairment and exacerbation (respectively) in their inotropic effects on LVDP, HR and RPP, when treating the HFD models and comparing them to the lean models (p<0.01 for each parameter for Dobutamine and p<0.0001 for each parameter for Itraconazole). PDH was significantly changed for the HFD hearts treated with Dobutamine (p<0.0001) and Itraconazole (p<0.0001), with little to no change in our atenolol treated hearts. UCP3 on the other hand showed no significance between the dobutamine and atenolol treated HFD and lean hearts but showed a significant decrease for the itraconazole HFD treated hearts when compared to the lean hearts (p<0.001). In conclusion, these results showed that both age and HFD cause functional changes in the heart in both absence and presence of drug treatments and their combined synergistic effect can cause further detrimental changes on different haemodynamic parameters. In addition, PDH activity, specifically, seems to be heavily linked to the inotropic effect of both Dobutamine and Itraconazole in the presence of obesity, but further work will still be necessary in order to confirm this observation.

Mathematical optimisation plays a crucial role in providing efficient solutions to modern engineering applications. Despite enormous advances in the past decades, some of these applications involve solving optimisation problems that may not be computationally tractable or that call for new theoretical advancements. Hence, obtaining robust and scalable methods and developing new analytic tools to optimisation problems is of paramount importance. This thesis addresses some of the main challenges when solving optimisation programs, namely, scalability, the presence of integer decision variables and the presence of uncertainty. Scalability throughout this thesis is achieved by means of distributed computation. Technological advancements that lead to the increase of computational devices per capita have attracted research on the development of algorithms that can exploit such a distributed and unstructured computational power to solve large-scale optimisation problems. In this context, multi-agent optimisation has emerged, as it enables distributed computation by allowing devices (or agents) to communicate over a network. This thesis proposes an algorithmic scheme based on subgradient averaging to perform multi-agent optimisation. Some optimisation problems are computationally intractable due to the presence of integer variables. Optimising over integers is, in general, hard due to the lack of polynomial-time algorithms to obtain an optimal solution. In fact, some well-known NP-hard problems can be cast as optimisation programs involving integer decision variables. In this thesis we investigate the so-called actuator placement problem and show that a particular instance of this problem, despite being an integer program can be solved exactly by means of a convex relaxation. We also link such a relaxation to the multi-agent optimisation framework explored previously, showing how distributed schemes can be leveraged to obtain the optimal solution. Other challenging optimisation problems arise with the presence of uncertain constraints. Indeed, even the well-studied class of linear optimisation problems may require theoretical and algorithmic advancements under this type of constraints. Uncertain constraints represent our lack of knowledge about the underlying phenomena and appear in several applications, including the optimal power flow problem under renewable energy generation, robotics, and transportation applications. Motivated by this fact, the last part of this thesis focuses on a randomised approximation to chance-constrained optimisation problems. We show a new theoretical bound on the probability of constraint violation that can improve the conservatism with respect to the state-of-the-art.

In this thesis, we study aspects of coding against synchronisation errors, such as deletions and replications, and related errors. Synchronisation errors are a source of fundamental open problems in information theory, because they introduce correlations between output symbols even when input symbols are independently distributed. We focus on random errors, and consider two complementary problems: We study the optimal rate of reliable information transmission through channels with synchronisation and related errors (the channel capacity). Unlike simpler error models, the capacity of such channels is unknown. We first consider the geometric sticky channel, which replicates input bits according to a geometric distribution. Previously, bounds on its capacity were known only via numerical methods, which do not aid our conceptual understanding of this quantity. We derive sharp analytical capacity upper bounds which approach, and sometimes surpass, numerical bounds. This opens the door to a mathematical treatment of its capacity. We consider also the geometric deletion channel, combining deletions and geometric replications. We derive analytical capacity upper bounds, and notably prove that the capacity is bounded away from the maximum when the deletion probability is small, meaning that this channel behaves differently than related well-studied channels in this regime. Finally, we adapt techniques developed to handle synchronisation errors to derive improved upper bounds and structural results on the capacity of the discrete-time Poisson channel, a model of optical communication. Motivated by portable DNA-based storage and trace reconstruction, we introduce and study the coded trace reconstruction problem, where the goal is to design efficiently encodable high-rate codes whose codewords can be efficiently reconstructed from few reads corrupted by deletions. Remarkably, we design such n-bit codes with rate 1-O(1/log n) that require exponentially fewer reads than average-case trace reconstruction algorithms.

Maps of Ixodes ricinus distribution and maps depicting tick bite risk are essential for government and health organisations to target prevention and control strategies for tick-borne diseases. However, researchers often lack robust long-term and geographically extensive tick distribution data, and information about human exposure to ticks to measure risk. Citizen science projects, through the collective effort of many volunteers, have the potential to provide valuable data on tick bite risk and tick distribution, but are often based on opportunistic submission of reports. The overarching aim of this study was to assess methodologies to improve public health decision-making through distribution mapping of ticks and tick bite risk for Scotland. Research was undertaken to: compare the quality and robustness of predictive mapping with the three types of tick data most often used for predictive mapping; use statistical approaches to improve the quality of predictions of the distribution of I. ricinus in Scotland, including the predicted uncertainty; assess whether questing tick surveys reflect human-tick encounter risk; and finally, test the feasibility of a new citizen science approach to assess human risk of tick encounters. Analysis of the three existing datasets with I. ricinus distribution in Scotland showed that whereas data from questing tick surveys generate detailed predictive maps at local scale, at the country level, predictions were affected by poor data coverage. Additionally, dissimilarities in the predicted distribution pattern of I. ricinus between data from passive submission and from questing tick surveys were identified. This suggests the need for data from public submissions to gather information on absences and to account for volunteer effort. A predictive map of I. ricinus distribution in Scotland developed with a sophisticated Bayesian statistical technique (the stochastic partial differential equation (SPDE)) which accounted for several sources of variation was successful in improving the predictions in areas with poor data coverage, and the associated uncertainty. The relationship between questing tick surveys and human tick bite rate was then assessed. Questing tick surveys were carried out whilst collecting contemporaneous data on tick encounters from orienteers running the same areas in 11 events at world, national, regional and local orienteering events. This novel approach found that questing tick surveys are a good indicator of tick bite risk. Also, the number of people multiplied by the hours of exposure is the most meaningful denominator for human exposure to tick bites (correlation coefficient with questing tick abundance of 0.8, p=0.0052). From 340 reports from orienteers recorded across all events, a mean incidence of 409 tick bites per 1,000 person-hours exposure was recorded. Significant correlations were found between tick bite rate and temperature on the event day, the proportion of pastures around the track used by orienteers and the start time of the activity. A citizen science project was implemented in Scotland between May and November 2018 and again between March and November 2019. The project used a novel approach that included collection of denominator data (number of people exposed, and time spent) and additionally asked people to report both when they did, and importantly, when they did not encounter ticks. Tick bite and tick encounter rates calculated from participant reports were compared with predictions of questing tick abundance in two study areas, Lochaber and the Cairngorms using data collected from questing tick surveys. A total of 1,914 reports from 65 volunteers were received, with 231 and 118 reports received, respectively, from the Cairngorms and Lochaber areas. On average, the Cairngorms area registered 0.083 tick bites per person per hour of activity and 0.268 tick encounters per person per hour. Lochaber area registered 0.018 tick bites and 0.028 tick encounters per person per hour. Tick bite and tick encounter rates in the Cairngorms correlated better with predicted tick abundance in the area (correlation coefficient of 0.27 and 0.31, respectively) compared to Lochaber (correlation coefficient of 0.15 in both cases). Tick bite and tick encounter rates were found to depend both on questing tick abundance, and on factors related to human activity and behaviour. Type of human activity explained more variation in tick bite rate than questing tick abundance. Tick bite and tick encounter rates were quantified by activity type. These findings are valuable in identifying high risk activities and targeting public health messages. This study resulted in new methodologies to improve predictive mapping of ticks, and better understanding of tick bite risk and the factors that drive it, with the overall aim of improving control and prevention of tick-borne diseases.

Higher education (HE)institutions play an important role in creating contexts for the development of identities. With respect to an increase in transnational education (TNE) within the Chinese HE environment over the last 10 to 15 years, the impact on individuals is under-researched. This thesis unpacks Chinese students' experiences in a TNE environment in China which, in turn, impacted on constructions of their cultural identities. The fieldwork included 31 semi-structured, arts-based, face-to-face interviews with students at Xi'an Jiaotong-Liverpool University (XJTLU), and is grounded in double-analytical lenses: Interpretative Phenomenological Analysis and Discourse Analysis. Choosing XJTLU was a complex culturally-driven process, and seemingly the second best alternative among other domestic choices. The transition period from a Chinese educational system to XJTLU significantly impacted on the construction of participants' personal and HE learner identities. Confucian values still form the fabric of Chinese social ethics, but new family values and relations are emerging. Articulation between students' personal and cultural values, and their embodied self and agency, suggested potentially conflicting forces between values related to culture, educational backgrounds, and desired identities. The relevance of TNE in China in shaping cultural identities of Chinese students relates to the importance of individual and cultural processes impacting on students' transforming self-perceptions and experiences. Students need to relate to themselves (as HE students) with: their peers, international academic staff, English as a medium of instruction use and how they feel they are perceived by others. This ongoing process of identity construction may have a profound impact on how these students see themselves as HE students and Chinese youth, and their place in the world into the future, and by extension on the way China as a nation will image itself in its engagement with the rest of the world, particularly in a post-COVID-19 era.

Safety testing procedures and risk assessment analysis of newly developed chemicals are mandatory to attest them as non-toxic for human handling. The European Union implemented several regulations enforcing safety testing and has recently banned the use of animals for toxicology studies for human safety. From in vitro cell-based assays to reconstructed human epidermis (RHE) models, there are several non-animal alternatives for toxicology studies to determine human safety. Cell-based assays are not accurate enough to predict the toxic potential of chemicals as they are a two-dimensional (2D) representation of skin cells. Three-dimensional (3D) RHE models are better, as they allow for differentiation of skin epidermal layers (i.e. stratum corneum). Commercially available RHE models are now the gold standard for toxicity studies for human safety and some of these have been validated by the Organization for Economic Cooperation and Development (OECD) for skin irritation and corrosion testing (two major categories of skin toxicity). The aim of this study was to develop an open source 3D epidermal skin culture based on Poumay's work. Using the Alvetex Strata scaffold, cell seeding was optimised using 1x106 keratinocytes for a 30 day culture period for best differentiation of the stratum corneum layer (confirmed by immunofluorescence staining of involucrin). Further optimization of vitamin C and calcium supplements did not seem to improve distribution of keratinocytes throughout the scaffold. Ultimately, the best results were achieved by the addition of a human-based collagen coat to the Alvetex Strata Scaffold, where development of stratum corneum was confirmed by collagen and involucrin staining. Moreover, development of the stratum basale layer was confirmed by cytokeratin 14 immunofluorescence. Evaluation of the 3D epidermal skin culture was tested following OECD Testing Guideline (TG) 439 for skin irritation in 12 reference chemicals from the 20 reference chemicals list in TG 439. Results showed the 3D epidermal skin culture was able to distinguish non-irritant from irritant chemicals using 10 different keratinocyte donors. Comparative performance analysis with a monolayer (2D) keratinocyte culture, a commercial RHE model and a collagen-based RHE models were performed. Results showed similar outcomes between the different models, while the monolayer culture performed poorly in discriminating irritant from non-irritant chemicals. Overall, these results were indicative of the predicative capacity of the 3D epidermal skin culture for in vitro testing of skin irritation. The second aim of this project was to use a skin explant assay, developed by Alcyomics, as a novel in vitro test for assessment of immunotoxicity caused by aggregation of monoclonal antibodies (mAbs). mAbs are important therapeutics but their potential for aggregation has become a critical quality parameter that can turn into a potential health risk during the administration of mAb therapeutics to patients. While the extent of immunotoxicity in patient populations is uncertain, reports show that it can lead to immune responses via cell activation and cytokine release. Our results showed that aggregated mAbs caused adverse immune events by evidence of tissue damage, expression of cell death markers and overall increase of IFN-γ (pro-inflammatory cytokine). These results showed that the skin explant assay could be a promising tool for predicating immunotoxicity caused by mAb aggregation. In conclusion, this study has provided further insight on the assessment of skin toxicity and immunotoxicity using in vitro skin assays.

This thesis explores trees as a vehicle for discussing and designing urban spaces. Green spaces and trees have been approached in their capacity to help cities cope with the urgent environmental challenges of climate change. Additionally, the unique performative qualities of trees – which include their spatial complexity – allow them to trigger processes of architectural and urban imagination. A multidisciplinary approach to trees, focusing on their roles in the city and in architecture, introduces spatial, ethical, cultural, environmental, and aesthetic modes of engagement. The design projects aim to capture specificities of trees in an array of urban contexts, from dense cityscapes to newly appropriated spaces and park landscapes. The exploratory approach to the spatial, narrative, cultural and aesthetic complexity of these landscapes allows me to ‘design from within’. The design methods, which unfold the spatial and architectural framework of the research, range from architectural drawings and performative exercises to terrestrial three-dimensional scanning technology. Informed by these modes, my heuristic process explores the articulation between natural elements and built structures in urban contexts. Ultimately, the thesis highlights the potential of terrestrial three-dimensional scanning in formulating original design responses that enable a speculative interpretation of trees in the city. The research explores both direct aspects of trees and the speculative potential of the urban landscapes in which they are integrated. Historical, analytical and associative accounts, alongside a fictional narrative, incorporate findings while immersing the reader into the ‘unconscious’ of the landscapes. The design process opens the path to an architecture that establishes spatial and metaphorical relationships with trees and the landscape. The multidimensional perspective on trees and their contexts alongside the original procedures of designing from within a landscape are an argument for a profound integration of the ‘natural’ and the ‘urban’ in the future design of cities.

Current treatments for macular degeneration, such as gene therapy, pharmacological approaches and neuroprotective approaches require the present of the target cells, the photoreceptors, in order to achieve a successful outcome. This leaves several patients, where the retinal degeneration is too advanced, without any viable therapeutic alternative. Photoreceptor transplantation aims to replace the lost cells, providing a potential option for such cases. For photoreceptor replacement to succeed, transplanted photoreceptors must mature and establish synaptic connection with the host retina. The remodeling of the host retina must be considered, as its interneurons and synaptic circuits rearrange following photoreceptor degeneration. Additionally, an ethical and renewable source is required. Studies have shown that mouse and human embryonic stem cells (ESC) can be differentiated to photoreceptor and transplanted into models of retinal dystrophy. However, no unambiguous evidence of synaptic connectivity and rescue of vision have been achieved. Here a detailed characterization of the remodeling events following photoreceptor lost in Aipl1-/- animals is described. Aipl1-/- were chosen due to the severe and fast emerging phenotype. As expected, typical features of remodeling were identified in these retinas and potential morphological elements were selected for analysis, following transplantation. The transplantation conditions, specifically the number of transplanted cells, were optimized using mouse ESC-derived photoreceptors. Here was no evident cell maturation or integration in the host’s synaptic circuit, following transplantation. Interestingly, when transplanting human (h) ESC-derived cones indication of such events was seen. To increase the period post-transplantation Aipl1-/- and Rd1, another well-established model of end-stage retinal degeneration, mice were crossed with immuno-compromised animals. Twelve weeks following transplantation into immuno-compromised Rd1 mice, hESC-derived cones matured and established functional synapses with the host retina, achieving rescue of vision. Alternative explanations for the rescue seen can be excluded due to the use non-functional human induced pluripotent (hiPS)-derived cones as a sham control.

Sodium benzoate (NaB) is a sodium salt that is widely used in the pharmaceutical, cosmetic and food industries. This widespread use results in almost everybody in the world being exposed daily to this compound. Currently very little is known about the effects of NaB can have on the immune response, even though it has been associated with the clinical course of chronic inflammatory diseases, such as orofacial granulomatosis and neurodegenerative disorders. Here, I show that THP-1 cells, a monocytic human cell line presents an altered immune response when exposed to NaB. Immunologically stimulated THP-1 cells in the presence of NaB secreted reduced levels of IL6 and IL1β and higher levels of TNF, while other cytokines such as IP10 and IL8 were unaffected. The inhibitory effect in IL6 and IL1β secretion was a consequence of a free radical scavenging characteristic of NaB, which neutralizes the reactive oxygen species (ROS) generated downstream of TLR activation. This resulted in the impairment of a secondary signalling event, which is required to fully activate the cells immune response. The use of microarray analysis in combination with q quantitative proteomic analysis revealed that NaB has a significant effect on the THP-1 cells beyond the alteration in cytokine secretion. NaB also interferes with cellular amino acid metabolism and has a major attenuating effect on the immune response. Taken together, these results suggest that NaB is not inert and has a major effect on a cells ability to mount an immune response. These findings could have major implications in how NaB is used in the future and in particular if it can be beneficial as a treatment for chronic inflammatory diseases, such as diabetes, atherosclerosis, rheumatoid arthritis, neurodegenerative disorders, and so on. On the other hand, by disturbing the inflammatory response, NaB could have a negative impact on other conditions such as orofacial granulomatosis. Further work will be needed to determine the role NaB plays in human inflammatory diseases.

This research work presents the design and implementation of a host-based Intrusion Detection and Prevention System (IDPS) called HIDROID (Host-based Intrusion Detection and protection system for andROID) for Android smartphones. It runs completely on the mobile device, with a minimal computation burden. It collects data in real-time, periodically sampling features that reflect the overall utilisation of scarce resources of a mobile device (e.g. CPU, memory, battery, bandwidth, etc.). The Detection Engine of HIDROID adopts an anomaly-based approach by exploiting statistical and machine learning algorithms. That is, it builds a data-driven model for benign behaviour and looks for the outliers considered as suspicious activities. Any observation failing to match this model triggers an alert and the preventive agent takes proper countermeasure(s) to minimise the risk. The key novel characteristic of the Detection Engine of HIDROID is the fact that it requires no malicious data for training or tuning. In fact, the Detection Engine implements the following two anomaly detection algorithms: a variation of K-Means algorithm with only one cluster and the univariate Gaussian algorithm. Experimental test results on a real device show that HIDROID is well able to learn and discriminate normal from anomalous behaviour, demonstrating a very promising detection accuracy of up to 0.91, while maintaining false positive rate below 0.03. Finally, it is noteworthy to mention that to the best of our knowledge, publicly available datasets representing benign and abnormal behaviour of Android smartphones do not exist. Thus, in the context of this research work, two new datasets were generated in order to evaluate HIDROID.

Cognitive brain function requires the establishment of neuronal networks, which rely on the formation and elongation and branching of axonal projections and the differentiation of presynaptic terminals during development. The cellular events involved in these processes are dependent on protein translation locally in the axon compartment, which enables rapid changes of the axonal proteome in response to neurotrophic cues to regulate axon growth and patterning. How these dynamic changes of the axonal proteome are regulated locally in the axon has been a topic of intensive investigation in the past decade. MicroRNAs are small RNAs known to regulate protein expression by controlling mRNA translation repression/degradation. Recently, these regulatory RNAs have emerged as key players in the modulation of several molecular pathways underlying neuronal differentiation in early stages of development, making this class of non-coding RNAs interesting candidate regulators of local protein translation in the axon during development. However, the role of microRNAs in the axon, in particular axonal outgrowth and presynaptic differentiation, is only beginning to be unravelled. The work described here used next generation sequencing to identify a set of microRNAs enriched in the axonal fraction of primary cortical neurons cultured in compartmentalised microfluidic devices. Following the characterization of axonal microRNA expression levels, two microRNAs, miR-3470b and miR-99a, were selected for subsequent functional studies. The miR-3470 family are mouse-specific repeat-derived microRNAs originating from the B1-element Mus1. Repetitive elements have a major role in shaping the structure and function of the genome, but evidence on the functional relevance of lineage-specific repeat-derived microRNAs is still limited. Inhibition of miR-3470b during early development of primary cortical neurons produced a significant decrease in axonal growth. Moreover, we discovered a significant association between miR-3470b targets and proteins involved in cell-to-cell contact/synaptic pathways. To investigate the role of miR-3470b in the formation of neuronal networks we used a microelectrode array cortical culture model, in which spontaneous electrical activity shows a progressive increase from day 9 in vitro, reflecting the establishment of functional synaptic connections. Inhibition of miR-3470b from the beginning of the third week in culture, when its endogenous levels are high, produced a marked decrease in network activity, suggesting its role in neuron connectivity during cortical neuron development. Functional studies in cortical cultures showed that inhibition of miR-99a in primary cortical neurons produced a significant decrease in axonal growth, with overexpression of a miR-99a mimic increasing axonal length. By using bioinformatics, luciferase reporter assays and functional rescue experiments we could identify a new target for miR-99a in axonal development, heparan sulphate 3-O-sulphotransferase 2 (Hs3st2), an enzyme part of the biosynthesis of heparan sulphate proteoglycans, key players in axon-extracellular matrix interactions. To investigate the effect of axonal miR-99a in the emergence of neuron connectivity and the functioning of neuronal networks we used imaging of calcium transients to assess the spontaneous rhythmic activity of developing neuronal cultures, demonstrating how inhibition of miR-99a alters the patterns of calcium oscillation frequency and synchronicity, suggesting its relevance for network development and maturation. Overall, this work identified two microRNAs capable of regulating axonal outgrowth in the development of mouse cortical neurons in vitro. Both microRNAs were found to exert growth promoting actions in developing axons. Furthermore, this work demonstrated the ability of mR-99a and miR-3470b to act as regulators of neuronal network formation in vitro, raising potential implications to the development of neuronal connectivity in vivo.

Chronic inflammation has long been considered an enabling characteristic of the cancer microenvironment. The release of growth and survival factors, extracellular matrix-modifying enzymes and other bioactive molecules by inflammatory cells into the tumour microenvironment contributes to the acquisition of cancer hallmark capabilities. It is now known that oncogenes can activate signals that promote the formation of an inflammatory microenvironment, even in the absence of external stimuli, placing inflammation as an early phenomenon in the timeline of tumour development. The over-expression of an oncogene in pre-neoplastic cells has been shown to promote the release of cytokines and other pro-inflammatory markers and consequent recruitment of neutrophils and macrophages. The recruitment of neutrophils has been previously associated with increased proliferation of oncogene-transformed cells, however, the mechanisms by which neutrophils exercise this trophic role are yet poorly understood. Nuclear factor kappa B (NF-kB) transcription factors are crucial elements for the regulation of inflammation, immune response as well as cellular stress response. Activation of NF-kB signalling pathway leads to the expression of mitogenic and anti-apoptotic factors, thus having the potential to promote tumorigenesis. MyD88 is an adaptor protein that mediates TLR/IL-1-R activation of the NF-kB pathway. MyD88 has been shown to play both positive and negative roles in cancer development. It is unclear however, whether the MyD88-NF-kB signalling pathway plays a role during the earliest stages of preneoplastic development of tumour initiation. To better understand the role of MyD88-NF-kB in oncogene driven inflammation during preneoplastic cell development, I was involved in the development of a zebrafish (Danio rerio) tissue specific inducible model which allowed the temporal control of HRasG12V oncogene expression for generation of pre-neoplastic cells (PNCs) and the detection of the earliest signalling events that initiate the process of tumour development. Using reporter lines, generated elsewhere, I demonstrate the activation of MyD88-NF-kB signalling pathway in both PNCs and recruited neutrophils. I also evaluate the effect of its downregulation, through the expression of the murine dominant negative mutant of IkBa, on PNCs proliferation and modulation of recruited neutrophil behaviour. Alongside the study of NF-kB as an inducer of tumour initiation, I also provide evidence of heterogeneity within the neutrophil population recruited to the PNCs and the different behaviours they exhibit. This work demonstrates the potential of zebrafish for the study of tumour initiation highlighting the involvement of NF-kB signalling pathway in the establishment of an inflammatory milieu that allows the progression of this first stage of tumorigenesis.

The ability to navigate through an environment and build mental representations of the space we visit, is an essential skill for many living creatures. In the hippocampus, space is represented through the activity of different spatial cell types, such as place cells, head-direction cells, and grid cells. In young animals, spatial cell types emerge at different ages: head-direction cells emerge in adult-like form at postnatal day (P) 16, but can be recorded before this age; grid cells emerge abruptly at P21; and place cells can first be recorded at age P16, but they gradually improve their firing characteristics. An also important spatial cell type, but often not talked about, are boundary encoding neurons in subiculum. Their physiology in the adult is well understood, but not much is known about their early postnatal features and how boundary information affects the firing of other spatial cell types. The work in this thesis aims to understand the postnatal development of subicular Boundary-Vector Cells (BVCs) as well as their functional relationship with entorhinal grid cells. The work here presented demonstrates that BVCs can be recorded from as early as P16, and like place cells, this spatial cell type gradually improves its firing characteristics until the animal reaches adulthood.

Advances in technology and incentives from governments are enabling growing deployment of distributed energy resources (DER) and this raises challenges over how distribution network operators (DNOs) should host these resources other than through traditional grid reinforcement. The alternative is active network management (ANM) which can control non-firm connections of DERs to ensure the network is compliant with voltage and thermal operating limits while increasing the capacity of the network to host DER. Furthermore, with changes expected in the regulatory paradigm to create ancillary service markets to promote other benefits of DERs, new streams of revenue for DER owners are anticipated. ANM schemes should manage growing volume of DERs while minimising costs of ancillary services. In this context, it is proposed and evaluated a distributed form of ANM for radial networks that manages voltage and thermal constraints and does this cost-effectively by including the costs of ancillary services in its objective function. To distribute the ANM, a mathematical formulation for local calculation of voltage sensitivities is derived that enables each local controller (LC) to take decisions over power injections to solve constraints. These estimates are combined in a message passing process that results in a coordinated set of decisions that solve local and non-local network constraint. The proposed ANM, based on estimated sensitivities and local message passing, is compared with a centralised ANM based on optimal power flow (OPF). Case-studies examined through time-domain simulations showed that the proposed ANM successfully identified the most appropriate DER to act and produced near identical control actions to the OPF. The approach is evaluated across multiple randomly generated scenarios, providing good confidence that the method is widely applicable and accurate. Mismatches of the order of 10^{-2} p.u where observed when compared to centralised algorithm and considerable decrease in computational time and data transmission was achieved.

Approximately 40% of the world's ice-free land has lost its original natural habitat cover to other land uses, such as croplands and pastures, which poses major challenges to conservation. Many conservation strategies have been developed to halt or compensate for past or future natural habitat loss, but the effectiveness of such strategies remains poorly understood. This thesis explores how an emerging conservation scheme can result in effective conservation gains. I use The Environmental Reserve Quota (or CRA; Portuguese acronym) within The Brazilian Forest Code as a case study - the largest forest offset scheme in the world. Under this scheme, landowners must offset past deforestation (buyers) by trading hectares of deforested land with landowners who have standing forest available in their properties (sellers). Alternatively, buyers can offset by allowing forest regrowth. By examining the viewpoints of potential buyers and sellers about CRA (Chapter 2), I find sharp divergences related to programme-specific factors that could potentially affect trade, mostly around price expectations and contract length. Next, I combine policy scenarios with potential spatial scales of scheme implementation, to test how scale affects scheme's conservation outcomes (e.g. area directed to avoided deforestation and/or regrowth; Chapter 3), as the scale in which conservation strategies are implemented often affect conservation trade-offs. Allowing offsets to occur within large spatial scales results in greater area of avoided deforestation and only a small area allocated to regrowth, whilst at small spatial scales results in the opposite pattern. However, the greatest total area was directed to conservation when the scheme was implemented at small scales. Finally, I compare the potential environmental co-benefits (e.g. aboveground biomass storage and accumulation and beta-diversity) associated with avoided deforestation and regrowth at large and small spatial scales (Chapter 4) and find that trade-offs between biomass and biodiversity co-benefits are dependent on spatial scale. Whilst large scale might protect biomass-rich forests, small scale protects highly threatened beta-diversity. These results are important for understanding important aspects associated to conservation schemes and their conservation outcomes. It is key that conservation polices account for landowners preferences in the design of schemes as well as the landscape scale in which conservation gains are expected to be delivered, in order to improve conservation effectiveness.

Bioprinting allows cells and materials to be deposited to create tissue pre-cursors. However, the quality of the bioprinted constructs depends on bio-inks, which need optimisation to enable repeatable processes, achieve physiological cell densities and tailored structures. The primary aim of this research was to develop reliable bio-inks for printing single cells and cells in gels. A novel bio-ink formulation has been developed to allow reliable inkjet bioprinting of small cell numbers, allowing single cell deposition. Poly-L-lysine (PLL) was investigated as temporary speckled coating, which would allow to avoid nozzle blockage during the printing process. Different PLL concentrations were evaluated, with 10 µg/mL exhibiting the highest viability and metabolic activity similar to control samples on three different cell lineages. At the optimised concentration, fast internalisation and metabolisation of PLL was observed with no subsequent effect on cell behaviour. Repeatable inkjet cell printing was achieved, whilst maintaining cell function. High cell density cell-laden hydrogels were produced using a bioprinting technique called Reactive Jet Impingement (ReJI), developed at Newcastle University as part of a related project. A protocol for using ReJI to produce a collagen-alginate-fibrin (CAF) gel was established and used to create cell-laden gels. The fabricated hydrogels exhibited rapid solution uptake and porous structures, both essential for cell migration. When culturing mesenchymal stem cells in the CAF gel a cell density of 40 million cells/ml supported faster osteogenic differentiation than a cell density of 4 million cells/ml. Additionally, increased cell density produced more organised tissue structures and increased expression of osteogenic biomarkers, indicative of accelerated bone tissue formation. This research has developed and characterised the performance of two bio-inks for two bioprinting techniques. It is concluded that careful consideration of the processing techniques and bio-ink formulation can allow for the creation of effective cell printing techniques for both single cells and cell-laden hydrogels.

A great deal of the research on cross-linguistic phonetic influence demonstrates that a speaker's knowledge of their first language (L1) significantly affects their ability to perceive and produce sounds in any other language. While current studies show that cross-linguistic transfer occurs at the L3 level, some research suggests that properties of both L1 and L2 are present in the production of L3 (Ionin, Montrul & Santos, 2011). Many studies have addressed perception, production and factors that influence foreign speech in Second Language Acquisition (SLA) (Watkins, Rauber & Baptista, 2009). As the number of multilingual individuals rises, so does the need for studies that investigate not only SLA but also that of additional languages (i.e., Third Language Acquisition). This dissertation examines how cross-linguistic influence (CLI) occurs among English, Spanish, and Brazilian Portuguese (BP), examining instances of vowel reduction, an aspect of phonological production. English and BP are assumed as vowel reducing languages, whereas Spanish displays negligible vowel reduction in comparison. The vowel productions in L3 BP of two multilingual groups, L1English-L2Spanish-L3BP (ESP) and L1 Spanish-L2 English-BP (SEP) were investigated in two tasks: a paragraph reading task (PRT) and a carrier phrase task (CPT). The study sought to determine whether i) a native speaker of a vowel reducing L1 and a non-vowel reducing L2 displays more or less vowel reduction in a vowel reducing L3 than a native speaker of a non-vowel reducing L1 and vowel reducing L2 and ii) how length of exposure to an L3 affects phonological production. Three fixed effects were considered: duration ratio, intensity ratio and height (F1). The goal was to ascertain whether the Typological Primacy Model (TPM) (Rothman 2011, 2015) or the L2 Status Factor Model (Bardel & Falk 2007, 2012; Hammarberg, 2001) would be a better predictor for how vowel reduction would occur in the L3. Results for duration ratio and vowel height showed no significant difference between groups ESP and SEP. Results for intensity ratio suggest L2 Status as a better predictor, as group SEP displayed more phonological transfer than the ESP group. A hybrid approach to L3 acquisition models is proposed. [The dissertation citations contained here are published with the permission of ProQuest LLC. Further reproduction is prohibited without permission. Copies of dissertations may be obtained by Telephone (800) 1-800-521-0600. Web page: http://www.proquest.com/en-US/products/dissertations/individuals.shtml.]

Statistical parametric speech synthesis (SPSS) has seen improvements over recent years, especially in terms of intelligibility. Synthetic speech is often clear and understandable, but it can also be bland and monotonous. Proper generation of natural speech prosody is still a largely unsolved problem. This is relevant especially in the context of expressive audiobook speech synthesis, where speech is expected to be fluid and captivating. In general, prosody can be seen as a layer that is superimposed on the segmental (phone) sequence. Listeners can perceive the same melody or rhythm in different utterances, and the same segmental sequence can be uttered with a different prosodic layer to convey a different message. For this reason, prosody is commonly accepted to be inherently suprasegmental. It is governed by longer units within the utterance (e.g. syllables, words, phrases) and beyond the utterance (e.g. discourse). However, common techniques for the modeling of speech prosody - and speech in general - operate mainly on very short intervals, either at the state or frame level, in both hidden Markov model (HMM) and deep neural network (DNN) based speech synthesis. This thesis presents contributions supporting the claim that stronger representations of suprasegmental variation are essential for the natural generation of fundamental frequency for statistical parametric speech synthesis. We conceptualize the problem by dividing it into three sub-problems: (1) representations of acoustic signals, (2) representations of linguistic contexts, and (3) the mapping of one representation to another. The contributions of this thesis provide novel methods and insights relating to these three sub-problems. In terms of sub-problem 1, we propose a multi-level representation of f0 using the continuous wavelet transform and the discrete cosine transform, as well as a wavelet-based decomposition strategy that is linguistically and perceptually motivated. In terms of sub-problem 2, we investigate additional linguistic features such as text-derived word embeddings and syllable bag-of-phones and we propose a novel method for learning word vector representations based on acoustic counts. Finally, considering sub-problem 3, insights are given regarding hierarchical models such as parallel and cascaded deep neural networks.

Automated analysis of medical imaging data allows both researchers and clinicians to develop more accurate and faster diagnoses. Image registration plays an essential role in both longitudinal studies and group analysis, allowing the combination of different imaging modalities, and automatic parcellation of regions of interest. Despite its wide use, image registration is still challenging with many issues such as artefacts, scarcity of correspondences, multi-modality, and computational complexity. Additionally, due to the lack of highly annotated datasets, the validation of image registration, and specifically non-linear registration, is also problematic. In this thesis several of these issues are addressed by introducing: a framework to validate non-linear registration methods; a robust and fast algorithm for non-linear registration; and validating the proposed methods in a conventional analysis. Current techniques used for non-linear image registration validation are explored, and it is shown that techniques based on label overlap are both not theoretically valid while also having poor accuracy. This analysis further leads to the development of a multiscale metric to minimize these problems. Also, a method based on the Demons Framework is proposed to improve the convergence speed of non-linear registration algorithms, and further extended to be robust in the presence of intensity inhomogeneities and contrast variations. The proposed methods are validated in a synthetic simulation platform with a known ground truth, compared with a manually traced region-of-interest, and tested in a voxel-based morphometry analysis of real data. It is shown that the proposed methods outperform other leading registration methods in both the synthetic simulation study and the manually traced data, and present reliable results in the voxel-based morphometry analysis. Furthermore, the impact of different registration algorithms is explored through the voxel-based morphometry study, and shown to affect the final results and their interpretation.

In this thesis calcite dissolution and precipitation are investigated during injection of CO2 WAG (water alternating gas) in limestone oil reservoirs. First, the equilibrium between calcite and the carbonic acid system is studied in a static environment to understand how variations in chemical composition, temperature and pressure affect the mineral reactions. Then, four models of CO2 solubility are presented (PHREEQC, CMG GEM, Duan & Sun and Diamond & Akinfiev) and compared against experimental data from the literature. An empirical model that couples the CO2 solubility to the mineral and aqueous reactions is constructed. After that, reactive transport simulations are performed using PHREEQC and GEM. The injection of carbonated water in a limestone reservoir is simulated with PHREEQC to assess the behaviour of calcite reactions. The obtained results are explained and also observed in a similar model using GEM. Additional simulations are performed in GEM concerning single-phase injection (seawater and pure CO2 injections) and their analyses are used to assist in the interpretation of the more complex CO2 WAG and CO2 SWAG (simultaneous water and gas injection). Different WAG slug sizes are simulated and simple relationships between the WAG ratio (volumetric ratio between injected water and injected gas at reservoir conditions) and the dissolved calcite are determined. Sensitivity of the porosity change and scale deposition is assessed during grid refinement. A dissolution zone around the injector wellbore is obtained for the WAG process that is dependent on the WAG scheme. Later, reactive transport simulations performed in GEM are extended to 3-phase non-isothermal flow in 2D and re-analysed. The impact of heat exchange on the dissolution is investigated for different time step sizes. More WAG scenarios are simulated and a relationship between the WAG scheme and dissolution similar to 1D simulations is obtained. The final model is completed by adding layers with different properties to investigate how the communication between layers affects the reactions. The dissolution zone and porosity increase are determined based on the flow capacities and front velocities of the layers, while the depth of more sever scale risks are located by considering also the gravity segregation of injected fluids. Finally, different geological scenarios, well operations and initial reservoir conditions are simulated.

Approximate verification frameworks are an approach to combat the well-known state-space explosion problem. For properties formulated as "no bad state can be reached", an approximate framework can be simply obtained by replacing exact reachability by some over-approximation. These frameworks look for a bad state in this over-approximation. If no bad states are found, the system satisfies the original property; the framework is sound. If a bad state is found, we have an inconclusive result: this bad state might be reachable or not. This permitted incompleteness is a cornerstone of such methods and it is a means to obtain efficiency. In this thesis, we propose three techniques to approximate reachability. The first analyses small subsystems to capture locally provable relationships between component states. The second and third complement the first by capturing some global system invariants. While the second combines component invariants to estimate whether components can cooperate to reach a system state, the third technique detects token mechanisms and deduces token invariants that approximate reachability. Moreover, we show how our approximations can give rise to useful frameworks that check deadlock freedom, static properties -- a class of state-based properties we propose -- and CSP's traces and failures refinement expressions. Our evaluation of these frameworks seems to suggest that they can efficiently prove a large spectrum of properties for a variety of systems. Despite tackling NP-hard problems, our frameworks can be efficiently powered by SAT/SMT solvers. The unsuccessful attempts at using symbolic exploration to precisely check concurrent systems, combined with the success of our frameworks, suggest that the approximations we study are a key factor in harnessing the verification power of these solvers.

The emergence of antibiotic resistance has been a serious concern for the last few decades. It hinders the treatment of infectious diseases, raising mortality and morbidity rates, as well as increasing the cost of healthcare. To investigate the problem of antibiotic resistant E. coli, 50 E. coli bacteraemia clinical isolates (Kent collection) were collected from East Kent Hospitals University NHS Foundation and phenotypically/genotypically characterized for antibiotic resistance, virulence factors, and the presence of putative plasmids. High levels of resistance were detected for amoxicillin and trimethoprim, and 14% of the isolates showed a multidrug-resistant phenotype. ST73 isolates exhibited the highest virulence potential while ST131 exhibited the highest levels of antibiotic resistance, although no correlation was detected between the two variables. In accordance with previous observations, co-carriage of CTX-M-15 and aac(3)-IIa, aac(6')Ib-cr, and blaOXA was observed in the collection, providing a possible explanation on why ESBL-producing isolates are often multidrug resistant. The production of nitric oxide (NO) by the mammalian immune system and its well-known anti-bacterial properties has prompted the investigation exogenously administered NO as an alternative to antibiotics. While combinatorial treatments of NO and antibiotics have proved to be successful against bacterial biofilms, this strategy has not been investigated in planktonic bacterial cells. Moreover, recent studies have shown that the generation of reactive oxygen species (ROS) resulting from hyperactivation of the aerobic respiratory chain of E. coli occurs in response to treatment with bactericidal antibiotics. This secondary effect of antibiotics is an important part of the as part of the lethality of bactericidal antibiotics under aerobic conditions. Given that NO is a well-known respiratory inhibitor, it was hypothesised this would diminish the toxic effects of antibiotics. To test this hypothesis, the effect of NO upon the lethality of a bactericidal antibiotic (gentamicin) was tested on a multidrug resistant E. coli. Pre-exposure to the NO-donor GSNO or NOC-12 prior to gentamicin treatment was found to increase bacterial tolerance to the antibiotic in planktonic cells: the presence of NO elicits a 10-fold increase in IC50 for gentamicin lethality against an E. coli clinical isolate. Further investigation showed that cytochrome bd-I, a NO-tolerant respiratory oxidase expressed maximally under microaerobic conditions in E. coli, is largely responsible for the sensitization of E. coli to gentamicin during NO exposure. The work herein reports that NO elicits a dramatic increase in the tolerance of E. coli to antibiotics. Hence, this work has revealed a huge void in knowledge related to 4 antibiotic potency during conditions relevant to infection (i.e. in the presence of NO). Furthermore, this work reveals that the cytochrome bd-I respiratory oxidase sensitises E. coli to antibiotics in the presence of NO. These findings shed light on how NO encountered during infection could impair the function of antibiotics and will prompt future research into how controlling levels of respiratory inhibition during infection may be used to improve antibiotic efficacy.

The costs of dams have been underestimated mainly due to the subjective matters involved in them. This practice-based research seeks to reveal the nature and the magnitude of these costs: the damage to people, livelihoods, communities, and, of course, to the environment. To investigate the perspectives of riverine people, as these are the peoples affected most by dam development, and to engage with their stories, this research blends methods and knowledge from the fields of Anthropology, Ecology, and Visual Arts in order to achieve its major aim: to communicate the intangible impacts caused by dams. Photography, in particular documentary photography, plays an important role in the way society operates and is shaped; this research explores this mode of communication with the view to empowering its subjects and giving greater voice to their stories. This project focuses specifically on hydropower schemes in Brazil. The method consisted of inviting people who have been affected by dam projects for hydropower purposes in three distinct areas of Brazil for an interview followed by a photo shoot in which they would be simultaneously the sitter and co-director. During the interview, every participant describes her/his story and feelings about the respective hydro project. The sitter then works collaboratively on the portrait, so that each image is based on an exchange between subject and researcher. The work, which comprises textual pieces that are in dialogue with these portraits and with other visual material gathered, represents how hydropower affects these people's lives as well as the riparian ecosystem. This practice is also analyzed within the frame of theories and insights concerning (i) photography in the political arena (like those of Ariella Azoulay, and Jacques Rancière), and (ii) perceptions of nature and development (like those of Eduardo Gudynas). This research considers and reaffirms transdisciplinary approaches as a refined means to access, understand, and present complex phenomena (like those of the dams and hydropower) as well as photography as an important agent in the processes of negotiation and transmission of knowledge between individuals, particularly those that involve intangible matters. It shows that affected people are important as voices to expose structures of power in societies, as well as advocating changes in our understanding of hydroelectricity, wealth, and welfare. It highlights that collaboration with the subject, and also with traditional communities, can play a vital role in the disclosure of sensitive knowledge and the way stories are told.

A capsule-based self-healing cementitious material, capable of autonomically repairing its own cracks, can extend the service life of concrete structures and decrease the costs associate with repair and maintenance actions. However, the size, shell thickness, shell material and mechanical properties of the capsules still need to be optimised to ensure self-healing performance. Thus, the objective of this research was to explore the controlled microfluidic encapsulation to investigate the production of microcapsules for physically triggered self-healing in cementitious materials. A flow-focusing microfluidic device was used to produce double emulsions to be selectively photopolymerised to generate a core-shell structure. Subsequently, the physical triggering was assessed by embedding the produced microcapsules in cement paste, fracturing it and observing the cracked surface in the SEM. The results showed the production of microcapsules with 80-140 μm of diameter with excellent control over size and shell thickness. Using water-in-oil-in-water (w/o/w) double emulsion, microcapsules were synthesised containing water, colloidal silica solution and sodium silicate solution as core material. In addition, an oil-in-oil-in-water (o/o/w) double emulsion was used to encapsulate mineral oil and emulsified healing agents. The formation of the core-shell structure with aqueous and organic cores was characterised using optical microscopy and SEM. It was demonstrated that the water is not retained inside of the capsule, resulting in the formation of dimples and buckled capsules, particularly for shells thickness ~7 μm. On the other hand, TGA confirmed the retention of mineral oil for shells thickness of ~2 μm and the encapsulation efficiency was demonstrated to be 66%. When the capsules were added to the cement paste, four key factors were observed to prevent physical triggering: (i) thick shells, (ii) buckling of thinner shells due to the loss of water core, (iii) mechanical properties and (iv) poor interfacial bonding. As a result, a mechanical characterisation of the shell material was performed, indicating brittle fracture at room temperature, reduced Young’s modulus when compared with cementitious matrix and stress at rupture of 15-36 MPa. In addition, an innovative methodology was proposed to functionalise the surface of the microcapsules with hydrophilic groups in order to increase the interfacial bonding between the cement paste and the microcapsules. Thus, microcapsules with low tensile strength, low shell thickness, organic core and good interfacial bonding were successfully synthesised and demonstrated to rupture upon crack formation. These results experimentally demonstrate the importance of reduced shell thickness, core retention and interfacial bonding as valuable guides during the design of microcapsules for physically triggered self-healing in cementitious materials.

Major changes in watermass chemistry during the opening of the central segment of the South Atlantic, and oceanic anoxic events (OAEs 1a to 1d), represent dramatic events that affected the ocean–atmosphere system during the Early Cretaceous. This thesis addresses the chemostratigraphic correlation of Tethyan and South Atlantic sections based on carbon-, sulphur- and strontium-isotopes (δ13C, δ34S and 87Sr/86Sr) integrated with other methods to explore the causal links between these phenomena. Firstly, the sedimentation of massive evaporite deposits is discussed and its relationship with the Early Aptian OAE 1a explored. The proposed coincidence in timing between massive evaporite deposition and OAE 1a enhances the understanding of carbon and sulphur cycles in the Early Cretaceous. Secondly, the top of sub-evaporite carbonates in the southern Brazilian basins is identified as having been deposited in a continental environment, and further evidence for its deposition during part of the OAE 1a interval is presented. Thirdly, it is suggested that the formation of the Ontong Java Plateau, already linked to the genesis of OAE 1a, promoted a eustatic sealevel rise responsible for rapid seawater entrance into the southern Brazilian basins during the Early Aptian. Fourthly, the first record of OAE 1a from northeastern Brazil (Sergipe- Alagoas Basin) has been recognized, allowing stratigraphic correlation with southern Brazilian basins. Stratigraphic correlation between the southern Brazilian basins is extended to the Falkland Plateau, where the decrease in the content of organic matter upwards in the stratigraphy of DSDP Site 511 through the Barremian–OAE 1b interval can be related to the evolution of South Atlantic basinal geometry. Finally, in the context of a Tethyan reference section, the record of the Barremian-Aptian boundary and the OAE 1a interval has been characterized in the Tethyan Poggio le Guaine scientific core from Marche–Umbria, Italy.

Single Nucleotide Polymorphisms (SNPs) are widely used molecular markers, and their use has increased massively since the inception of Next-Generation Sequencing (NGS) technologies, which allow detection of large numbers of SNPs at low cost. However, both NGS data and their analysis are error-prone, which can lead to the generation of false positive (FP) SNPs. The traditional approach to SNP discovery is based on mapping reads to a reference sequence. Apart from sequencing errors, which vary in pattern and rate depending on the sequencing platform, the short read lengths that prevail in NGS, together with the repetitive nature of the genomes of many organisms, can lead to errors in the genome assembly and/or read mapping stages of the mapping-based approach for SNP discovery. The work described here has investigated and quantified some mechanisms that cause false positive SNPs. These include reference misassembly due to the presence of paralogous sequences and read cross-mapping, along with associated factors such as quality of the reference sequence, read length, choice of mapper and variant caller, mapping stringency, and filtering of SNPs by read mapping quality and read depth. The study shows that both paralogs and the choice of tools and parameters involved in variant calling can have a dramatic effect on the number of FP SNPs produced. A brief exploration of the influence of these factors towards false negative (FN) SNPs generation is also carried out in the end of the study, paving the way to new insights. This thesis aims to provide a stepping stone towards a better understanding of the factors influencing the mapping-based SNP discovery approach.

Many exchange rate papers articulate the view that instabilities constitute a major impediment to exchange rate predictability. In this thesis we implement Bayesian and other techniques to account for such instabilities, and examine some of the main obstacles to exchange rate models' predictive ability. We first consider in Chapter 2 a time-varying parameter model in which fluctuations in exchange rates are related to short-term nominal interest rates ensuing from monetary policy rules, such as Taylor rules. Unlike the existing exchange rate studies, the parameters of our Taylor rules are allowed to change over time, in light of the widespread evidence of shifts in fundamentals - for example in the aftermath of the Global Financial Crisis. Focusing on quarterly data frequency from the crisis, we detect forecast improvements upon a random walk (RW) benchmark for at least half, and for as many as seven out of 10, of the currencies considered. Results are stronger when we allow the time-varying parameters of the Taylor rules to differ between countries. In Chapter 3 we look closely at the role of time-variation in parameters and other sources of uncertainty in hindering exchange rate models' predictive power. We apply a Bayesian setup that incorporates the notion that the relevant set of exchange rate determinants and their corresponding coefficients, change over time. Using statistical and economic measures of performance, we first find that predictive models which allow for sudden, rather than smooth, changes in the coefficients yield significant forecast improvements and economic gains at horizons beyond 1-month. At shorter horizons, however, our methods fail to forecast better than the RW. And we identify uncertainty in coefficients' estimation and uncertainty about the precise degree of coefficients variability to incorporate in the models, as the main factors obstructing predictive ability. Chapter 4 focus on the problem of the time-varying predictive ability of economic fundamentals for exchange rates. It uses bootstrap-based methods to uncover the time-specific conditioning information for predicting fluctuations in exchange rates. Employing several metrics for statistical and economic evaluation of forecasting performance, we find that our approach based on pre-selecting and validating fundamentals across bootstrap replications generates more accurate forecasts than the RW. The approach, known as bumping, robustly reveals parsimonious models with out-of-sample predictive power at 1-month horizon; and outperforms alternative methods, including Bayesian, bagging, and standard forecast combinations. Chapter 5 exploits the predictive content of daily commodity prices for monthly commodity-currency exchange rates. It builds on the idea that the effect of daily commodity price fluctuations on commodity currencies is short-lived, and therefore harder to pin down at low frequencies. Using MIxed DAta Sampling (MIDAS) models, and Bayesian estimation methods to account for time-variation in predictive ability, the chapter demonstrates the usefulness of suitably exploiting such short-lived effects in improving exchange rate forecasts. It further shows that the usual low-frequency predictors, such as money supplies and interest rates differentials, typically receive little support from the data at monthly frequency, whereas MIDAS models featuring daily commodity prices are highly likely. The chapter also introduces the random walk Metropolis-Hastings technique as a new tool to estimate MIDAS regressions.

This thesis focuses on theoretical quantum optics, with special emphasis on developing protocols in which engineered vacuum forces enable one to construct hybrid systems. In these systems, atoms are combined with solid-state devices in order to take advantage of their unique properties such as long coherence times of atoms and flexibility, tunability, scalability, and fast response offered by solid-state systems. Special attention is given to the study of atom-surface interactions with Rydberg atoms, where exact Fano-type diagonalization of the interaction Hamiltonian is obtained showing that, not only do Rydberg atoms suffer energy shifts, the presence of a surface leads to an alteration and admixture of the unperturbed eigenstates. Of particular interest are dispersion forces on graphene systems. We investigate whether and under which circumstances the Casimir-Polder potential between an atom and a graphene-substrate system is dominated by the interaction with graphene such that the effect of the substrate does not play an important role. We also explore the possibility to create a setup where dispersion forces could be use to bend a graphene sheet. Placing an atom close, at distances of a few hundred nanometers, to a free-standing graphene membrane we show that temporal changes in the atomic state change the Casimir-Polder interaction, thereby leading to the creation of a backaction force in the graphene sheet. Finally, we look at nonlinear atom--surface coupling processes with the aim of proposing a hybrid quantum circuit device in which individual field-excitations can be transferred between atoms and surface polaritons on demand. Deeper investigations of nonlinear processes reveal the existence of a sum rule for two-photon spontaneous decay rates that can be simply understood as a redistribution of photonic modes across the frequency spectrum where the total integrated number of modes is still conserved.

Ultrasound is a commonly used imaging modality in diagnosis and pre-operative assessment of breast masses. However, radiologists often find it very difficult to correctly size masses using conventional ultrasound images. Consequently, there exists a strong need for more accurate sizing tools to avoid either the removal of an over-estimated amount of tissue or a second surgical procedure to remove margins involved by tumour not removed in the primary operation. In this thesis, we propose a new method of processing the backscattered ultrasound signals from breast tissue (based on the Fourier spectral analysis) to better estimate the degree of echogenicity and generate parametric images where the visibility of breast mass boundaries is improved (SPV parametric image). Moreover, an algorithm is proposed to recover some anatomical structures (particularly, Cooper’s ligaments) which are shadowed during the image acquisition process (LWSPV parametric image). The information from both algorithms is combined to generate a final SPV+LWSPV parametric image. A 20-case pilot study was conducted on clinical data, which showed that the SPV+LWSPV parametric image added useful information to the B-mode image for clinical assessment in 85% of the cases (increase in diagnostic confidence in at least one boundary). Moreover, in 35% of the cases, the SPV+LWSPV parametric image provided a better definition of the entire boundary. Note that the radiologist knew the final diagnosis from histopathology. In addition, the SPV+LWSPV method has the advantage that it uses the I/Q data from a standard ultrasound equipment without the need for additional hardware. On the basis of these facts, we believe there to be a case for further investigation of the SPV+LWSPV imaging as a useful clinical tool in the pre-operative assessment of breast mass boundaries.

In the last decade unprecedented improvement in cure rates and overall survival was achieved in diffuse large B-cell Lymphoma (DLBCL) through the introduction of rituximab and anthracyclin-based chemotherapy (R-CHOP) as first line treatment. However, 40% of patients are refractory or relapse after R-CHOP and are hardly salvaged. To date, only age, International Prognostic Index (IPI) stratification and genetic aberrations defining gray-zone lymphomas have been used in clinical trials to select high-risk patients for more aggressive regimens. However, these prognostic features do not take into account the full biological heterogeneity of DLBCL. This reflects our limited knowledge on comprehensive prognostication in this group of disorders and supports our choice to investigate old and new prognostic factors for DLBCL in this thesis. Molecular characterization is generating opportunities for personalized therapy in poor-risk DLBCL. In order for targeted therapies to succeed in this disease, reliable and reproducible strategies that adequately segregate patients into distinct molecular groups are needed. While gene expression profiling (GEP) is the gold standard method, there is presently a lack of standardized methodology for array analysis, which can lead to variable results. The lack of a routine methodology for GEP has led investigators to develop immunohistochemistry (IHC) based approaches for the molecular classification in DLBCL. In fact, the Hans algorithm is being used to identify non-GCB DLBCLs in clinical trials offering NF-kB targeting agents to patients with this subtype. By performing a systematic comparison of nine IHC algorithms for molecular classification in a new large dataset of diagnostic DLBCL, we document an extremely low concordance across all classifiers (<21%) when classifying each individual patient, and a lack of outcome impact of all strategies, demonstrating that IHC is not a reliable alternative to molecular-based methods to be used for clinical decisions in DLBCL. GEP studies also suggested that the microenvironment could provide prognostic biomarkers in DLBCL in the R-CHOP era. Most authors have focused on the use of IHC to enumerate and functionally characterize the microenvironment in DLBCL. In our second study, by comparing two methods of semi-automated analysis for IHC staining Abstract 6 of the microenvironment, we demonstrate that the computerized results are highly reproducible, add the required robustness to IHC studies and should be used in the future instead of manual analysis. By applying comprehensive statistical analysis we propose that CD3 and FoxP3 should be validated as predictors of response to R-CHOP in clinical trials. Whereas a number of mechanisms by which cancer cells influence macrophage function have been described, currently there is very limited understanding of the macrophage polarisation status and effector function in human DLBCL. In our third study we analysed the GEP of macrophages sorted from human DLBCL samples. Unsupervised hierarchical clustering does not resolve DLBCL macrophage samples from reactive macrophage samples, indicating that macrophage heterogeneity in DLBCL should be considered. 202 genes are differentially expressed in DLBCL relative to controls. Functional annotation supports that these genes are macrophage-specific. We demonstrate that DLBCL macrophages have a bidirectional M1 and M2 functional activation, challenging the concept, widespread in the literature, that macrophages in tumours have a predominant M2 transcriptome. In our fifth study we used a two-cell co-culture model in an attempt to demonstrate that DLBCL cells influence macrophage transcriptome and proteome. The heterogeneity of the results, which precludes the confirmation of our hypothesis, is fully discussed. In our last study we tease out the DLBCL macrophage GEP heterogeneity and propose IFN- as a culprit B-cell derived molecule influencing macrophage activation status. Finally, using immunofluorescence we demonstrate that both M1 and M2 proteins are expressed in DLBCL macrophages.

In the formal modelling of systems, demonic and angelic nondeterminism play fundamental roles as abstraction mechanisms. The angelic nature of a choice pertains to the property of avoiding failure whenever possible. As a concept, angelic choice first appeared in automata theory and Turing machines, where it can be implemented via backtracking. It has traditionally been studied in the refinement calculus, and has proved to be useful in a variety of applications and refinement techniques. Recently it has been studied within relational, multirelational and higher-order models. It has been employed for modelling user interactions, game-like scenarios, theorem proving tactics, constraint satisfaction problems and control systems. When the formal modelling of state-rich reactive systems is considered, it only seems natural that both types of nondeterministic choice should be considered. However, despite several treatments of angelic nondeterminism in the context of process algebras, namely Communicating Sequential Processes, the counterpart to the angelic choice of the refinement calculus has been elusive. In this thesis, we develop a semantics in the relational setting of Hoare and He's Unifying Theories of Programming that enables the characterisation of angelic nondeterminism in CSP. Since CSP processes are given semantics in the UTP via designs, that is, pre and postcondition pairs, we first introduce a theory of angelic designs, and an isomorphic multirelational model, that is suitable for characterising processes. We then develop a theory of reactive angelic designs by enforcing the healthiness conditions of CSP. Finally, by introducing a notion of divergence that can undo the history of events, we obtain a model where angelic choice avoids divergence. This lays the foundation for a process algebra with both nondeterministic constructs, where existing and novel abstract modelling approaches can be considered. The UTP basis of our work makes it applicable in the wider context of reactive systems.

Heterozygous mutations in the glucocerebrosidase (GBA) gene represent the most common risk factor for Parkinson’s disease (PD), a disease in which midbrain dopaminergic neurons are preferentially vulnerable. However, the mechanisms underlying this association are still unknown, mostly due to the lack of an appropriate model of study. In this thesis, we aimed at elucidating the role of heterozygous GBA mutations in PD using a specific human induced pluripotent stem cell (hiPSC)-based model of disease. First we developed a protocol for the efficient differentiation of hiPSCs into dopaminergic cultures, and extensively characterized the derived dopaminergic neurons which expressed multiple midbrain relevant markers and produced dopamine. Next we screened a clinical cohort of PD patients to identify carriers of GBA mutations of interest. Using for the first time hiPSCs generated from PD patients heterozygous for a GBA mutation (together with idiopathic cases and control individuals) we were able to efficiently derive dopaminergic cultures and identify relevant disease mechanisms. Upon differentiation into dopaminergic neuronal cultures, we observed retention of mutant glucocerebrosidase (GCase) protein in the endoplasmic reticulum (ER) with no change in protein levels, leading to upregulation of ER stress machinery and resulting in increased autophagic demand. At the lysosomal level, we found a reduction of GCase activity in dopaminergic neuronal cultures, and the enlargement of the lysosomal compartment in identified dopaminergic neurons suggesting a decreased capacity for protein clearance. Together, these perturbations of cellular homeostasis resulted in increased release of α-synuclein and could likely represent critical early cellular phenotypes of Parkinson's disease and explain the high risk of heterozygous GBA mutations for PD.

Affective engineering applies mathematical models to convert the information obtained from persons’ feelings to product features into an ergonomic design. However, the methods commonly used to elicit persons’ responses can present inaccuracies if measurement principles are violated. Consequently, empirical studies cannot be easily replicated and results cannot reliably be compared. This research aimed to overcome the problem by establishing a novel approach in affective engineering using probabilistic models underpinned by Rasch measurement theory. The Rasch model verifies whether the observations meet the assumptions necessary for quantifying the numerical validity of the data employing the tools of standard statistics. Initially, the research examined how well the data from affective responses would fit the expectations of the Rasch model to create a scale of specialness for four pieces of wrapped confectionery. Anomalies in the data were investigated to determine their potential impact on measure interpretation. A second empirical study investigated the stability of a measurement structure. Affective responses were compared with the physical properties related to compliance of a collection of product containers. A cross-validation strategy contrasted calibrations of the scale using different groups of respondents. The results indicated that the differences between person locations on the measurement continuum from different calibrations were statistically non-significant. This provided evidence that the use of a Rasch-calibrated scale can systematically refine and generalise its frame of reference without loss of measurement properties. The contribution of the research for the advancement of knowledge is established by transforming affective responses to physical elements into objective measures. A rationale was developed to achieve measurement properties in scales used in affective engineering, adapting the Rasch model’s taxonomy used in other domains. Furthermore, the stability of a scale using different samples for calibration is demonstrated to be a property of Rasch-based scales. As a consequence of the stability, the association between affective responses and sensory information was realised and further variables were incorporated in the calibrated metric to refine the understanding of users’ experience. Consequently, reliable results can be obtained from small samples, which will reduce time and costs of quantitative consumer research.

The aim of this project is to investigate the suitability of specific polymers inclusions in improving some important mechanical properties of single crystals of calcite. The mechanical properties of synthetic and mineral single crystals of pure calcite are reported and compared with synthetic and biogenic crystals incorporating polystyrene particles or di-block co-polymer micelles and organic molecules respectively. The mechanical properties of these crystals are measured using nanoindentation and micro-compression techniques. It is shown that the crystals’mechanical properties obtained by the nanoindentation Oliver-Pharr method are influenced by the compliance of the resin substrate into which the crystal samples are embedded. The embedding of the crystals is shown to be necessary in order to achieve a flat surface onto which reliable indentation can be made. An estimate of the elastic moduli of the samples was obtained by using the Song-Pharr model, which takes into account substrate compliance. The hardness of all the tested samples was also recalculated by using conventional methods. By analyzing the obtained substrate independent data it was found that although the occlusion of co-polymer micelles reduced the modulus of pure calcite its hardness was not affected. This behaviour was not observed in the composites occluded with 200 nm polystyrene particles which showed a significant hardness reduction. The 20 nm co-polymer micelles also increased the specific hardness of the crystals whereas the 200 nm polystyrene particles showed no such behaviour. The micro-compression data showed that the 200 nm polystyrene particles could potentially be used to reduce the brittleness of calcite. Images of cracks, post-fracture of the crystals, indicated deformation of the polymer. This is evidence of the occurrence of a crack bridging mechanism. It was thought that this could lead to an enhancement of the strain at fracture and work of fracture of these composites when compared with pure calcite. Mechanical data from these tests showed however that the polystyrene particles’ occlusion was inadequate to increase the strain to fracture and the work of fracture of calcite. In addition to this the specific compressive strength, specific compressive modulus and specific compressive work of fracture of calcite also decreased with the addition of PS particles.

This thesis examines how scientific communities which are heterogeneous among themselves communicate and collaborate to produce knowledge on climate change. Climate-change science is a relatively new field of investigation and it includes experts from virtually all areas of scientific enquiry. This field is not, however, a homogeneous transdisciplinary area of research so that the different scientific communities that compose it have to bridge the gaps among themselves to be able effectively to communicate and collaborate. I use Collins and Evans’ (2007) realist theory of expertise combined with other relevant Science and Technology Studies concepts, particularly the notions of trading zones (Galison 1997; Collins et al. 2007) and trust (Giddens 1990, Shackley and Wynne 1995b; Reyes-Galindo 2011) to explain how different groups of experts build bridges between their heterogeneous forms of life. As climate-change science is too broad to be covered in one PhD project, I focus on paleoceanography, a subfield of geology that reconstructs past oceans and their interactions with the climate system. I use the fractal model (Collins 2011) to move through different levels of analysis and examine the different bridge-building mechanisms between expert communities at work at each of them. The main contribution of the present work is to identify and explain how the various mechanisms that mediate communication between expert communities come into play at different levels of analysis.

Neurofibromatosis type 1 (NF1) is a common genetic disorder that predisposes to the development of heterogeneous tumours of Schwann cell origin, termed neurofibromas. Neurofibromas are thought to arise from a combination of genetic events – loss of the Ras-GAP neurofibromin in the Schwann cell lineage – and microenvironmental cues. Schwann cells are specialised cells that ensheath and myelinate the axons in the peripheral nervous system (PNS). In the adult they are present in a quiescent state, however following damage to the PNS they have a remarkable ability to regenerate. Distal to the site of injury, Schwann cells dedifferentiate to a progenitor-like state, in which they contribute to nerve repair by recruiting a robust inflammatory response and helping axons return to their targets. Work from our laboratory has shown that activation of the Ras/Raf/ERK pathway plays a central role in driving the switch in Schwann cell state from a fully differentiated to a proliferating, “progenitor-like” cell. Crucially, neurofibromas resemble injured nerves in that they are composed of a mixture of inflammatory cells and Schwann cells that are found dedifferentiated and dissociated from axons, suggesting that deregulation of Ras/ERK may trigger tumourigenic events. In this thesis I present work on how the Ras/Raf/ERK pathway may be regulated in Schwann cells. I show that the phosphatase MKP3 may be involved in controlling the levels of ERK activity in Schwann cells during differentiation and following nerve injury. I also describe a new model for neurofibroma formation. Using transgenic mice I show that Nf1 loss in adult, myelinating Schwann cells has no effect on peripheral nerves and does not induce tumourigenesis. However, when coupled with an injury, the mice developed tumours at a high frequency. Furthermore, I show that in the absence of Nf1, ERK signalling is deregulated upon injury, implicating this pathway in the tumour formation. This may have therapeutic relevance, which is currently being tested in our animal model. In addition, we observed that tumours only arise at the wound site, despite Schwann cells dedifferentiating along the length of the nerve. This strongly implies that the microenvironment is a crucial player in the outcome of Nf1 loss and reveals this new animal model as a promising system to further dissect molecular events involved in tumourigenesis.

The formation of the central nervous system depends on the patterned generation of different neuronal and glial cell types during embryogenesis. In ventral regions of the neural tube, the graded activity of the secreted molecule Sonic Hedgehog (Shh) controls the position along the dorsoventral axis at which different neuronal subtypes differentiate. Here, using chick and mouse embryos, we systematically and quantitatively documented the spatial and temporal expression patterns of the transcription factors that delineate the progenitor domains of these neurons. We investigated the coordination of the patterning activity of Shh with tissue growth and cell differentiation and discovered a high degree of conservation in pattern formation between mouse and chick. In addition, quanti"cation of the levels of Shh protein in vivo and downstream signalling activity using a mouse reporter of Gli activity con"rmed the non-linear relation between ligand concentration, levels of intracellular signalling and expression of the target genes in vivo. These data support a model in which progenitor cells respond to the cumulative amount of Gli activity. To advance our understanding of how cells transduce the Shh signal we investigated the vertebrate orthologs of Costal2, a component of the Hedgehog pathway in Drosophila. Using a gain of function assays in chick we con"rmed that one of the orthologs, the kinesin Kif7, is a negative regulator of the Shh signalling pathway, acting downstream of Smoothened. The inhibitory activity of Kif7 required only the N-terminal region containing the motor domain. By contrast, the kinesin Kif27, which appears equally similar to Costal2 as Kif7, was unable to inhibit Shh signalling. These data provided evidence that Kif7 was a novel component of the vertebrate Hh signalling and revealed a greater conservation between the Drosophila and vertebrate system than previous views suggested.

Knowledge is a powerful tool, especially within companies and institutions. It has long been recognised that much of the knowledge of a company is embedded in the skills and practices of individuals and groups of individuals that share common interests. In some cases, these groupings can cross traditional organizational boundaries. These communities, known as Communities of Practice (CoPs), have a significant potential for companies and institutions. Although studied for years now, Communities of Practice are still poorly understood. Specifically, there has been little research into the dynamic, temporal aspects of how communities form, their patterns of activity and how they cease to exist. Such understanding is crucial for the successful use of CoPs within companies and institutions. Therefore, this thesis looks at the temporal aspects of Communities of Practice (CoPs). This is carried out by studying a co-located CoP, and four CoPs located within electronic networks. The study led to a rediscovery of the already known concept of CoPs’ core members, which refers to the members with strong participation in the community’s life. This work also identified two completely novel aspects related to the temporal aspects of Communities of Practice. The first one was called Recurrent Communities of Practice (RCoPs). This refers to a new class of CoPs found inside the studied co-located Community of Practice (CoP). Recurrent CoPs are Communities of Practice that work over a specific period of time, ceasing their activity until a trigger starts a new period, repeating these two states over time. The Recurrent CoPs usually have the same participants in different periods of activity. The second new concept was named Transient Core Members (TCMs). This refers to a community’s isolated members that do not engage in the community’s activity constantly, but rather in “bursts” over time, similarly to RCoPs. It is hoped that the procedures and methods explained in this work might help the development for new tools for researchers and professionals to detect and nurture Communities of Practice (CoPs) and Recurrent Communities of Practice (RCoPs). Specifically, the research expands our understanding of the temporal aspects of Communities of Practice, seldom discussed previously in the area.

The central subject of this thesis is the application of the concepts of symmetry and proportion in music and how these can be used to generate original compositions. Information about the musical application regarding concepts of symmetry and proportion both in the twentieth-century and earlier is provided. The first section also offers additional information about the construction of symmetrical harmony and its common usage; here too, principles of intervallic proportion are explained based on the compositional thinking of English composer Christopher Bochmann. The second section presents seven original compositions, each supported by their own commentary. Each work features a variety of instrumental forces ranging from solo to orchestral. Lending a separate emphasis (in analysis) to each composition helps to provide a broad picture of the potential that the ideas of symmetry and proportion bring to contemporary composition.

The perennial controversy in business combinations accounting and its dialectic with stakeholders’ interests under the complexity of the Mergers and Acquisitions (M&A) activity is the centrepiece of analysis in this thesis. It is argued here that the accounting regulation should be as neutral as possible for the economic activity, although it is recognised that accounting changes may result in economic effects. In the case of the changes for business combinations accounting in the USA, lobbying was so fierce that in order to achieve the abolition of accounting choice in M&A accounting, it forced the standard-setter to compromise and to change substantially some of its earlier proposals. Such fierce lobbying cast doubts about whether it was effectively possible to mitigate such economic effects, resulting in a possible impact of the accounting changes on the M&A activity. The occurrence of M&A in waves is yet to be fully theorised. Nevertheless, existing literature established relationships between M&A activity and some key economic and financial factors, and has provided several interesting theories and other meaningful contributions for this thesis. It was therefore possible to examine whether the changes in the accounting rules produced any significant impact on the M&A activity. The findings obtained from the testing of the research hypotheses suggest that the new M&A accounting rules did not result in significant impacts on overall M&A activity. Nevertheless, from the study of managers’ perceptions, and from the examination of annual reports of S&P 500 companies, a considerable impact on the financial reporting was found.

Mitosis is the shortest phase of the cell cycle but visually the most outstanding. The key goal of mitosis is to accurately drive chromosome segregation. On one hand, DNA has to be condensed into characteristically shaped chromosomes. On the other hand, a very specialized structure needs to be built to conduct segregation, the mitotic spindle which is composed of microtubules organized into an antiparallel array between the two poles. The interaction between microtubules and chromosomes occurs at the kinetochore, a macromolecular complex assembled in mitosis at the centromere. The centromere/kinetochore monitors proper spindle microtubule attachment to each of the chromosomes, aligning them at the metaphase plate and also ensuring that chromosome segregation happens in perfect synchrony. Although centromeres are present in all eukaryotes, their basic structure and chromatin folding are still poorly understood. One of the aims of my work was to understand the function of the condensin complex specifically at the centromere during mitosis. Condensin I and II are pentameric protein complexes that are among the most abundant components of mitotic chromosomes. I have shown that condensin is important to confer stiffness to the innercentromeric chromatin once spindle microtubules interact with kinetochores in metaphase. Labile inner-centromeric regions delay mitotic progression by altering microtubule-kinetochore attachments and/or dynamics with a consequent increase in levels of Mad2 checkpoint protein bound to kinetochores. In the absence of condensin, kinetochores perform prominent “excursions” toward the poles trailing behind a thin thread of chromatin. These excursions are reversible suggesting that the centromeric chromatin behaves like an elastic polymer. During these excursions I noticed that only the inner centromeric chromatin was subjected to reversible deformations while the kinetochores (inner and outer plates) remained mostly unaltered. This suggested that the centromeric chromatin part of the inner kinetochore plate was organised differently from the subjacent chromatin. I went on to investigate how the centromeric chromatin is organised within the inner kinetochore domain. Super-resolution analyses of artificially unfolded centromeric chromatin revealed novel details of the vertebrate inner kinetochore domain. All together, the data allowed me to propose a new model for the centromeric chromatin folding: CENP-A domains are interspersed with H3 domains arranged in a linear segment that forms planar sinusoidal waves distributed in several layers. Both CENP-A and H3 arrays face the external surface, building a platform for CCAN proteins. CENP-C binds to more internal CENP-A blocks thereby crosslinking the layers. This organization of the chromatin explains the localisation and similar compliant behaviour that CENP-A and CENP-C showed when kinetochores come under tension. Other kinetochore proteins (the KMN complex) assemble in mitosis on top of the CCAN and bind microtubules. KMN binding may confer an extra degree of stability to the kinetochore by crosslinking CENP-C either directly or indirectly. My work and the testable model that I have developed for kinetochore organization provide a fundamental advance in our understanding of this specialized chromosomal substructure.