Outlier effects on robust joint modelling of longitudinal and survival data

Robust joint modelling is an emerging field of research. Through the advancements in electronic patient healthcare records, the popularly of joint modelling approaches has grown rapidly in recent years providing simultaneous analysis of longitudinal and survival data. This research advances previous work through the development of a novel robust joint modelling methodology for one of the most common types of standard joint models, that which links a linear mixed model with a Cox proportional hazards model. Through t-distributional assumptions, longitudinal outliers are accommodated with their detrimental impact being down weighed and thus providing more efficient and reliable estimates. The robust joint modelling technique and its major benefits are showcased through the analysis of Northern Irish end stage renal disease patients. With an ageing population and growing prevalence of chronic kidney disease within the United Kingdom, there is a pressing demand to investigate the detrimental relationship between the changing haemoglobin levels of haemodialysis patients and their survival. As outliers within the NI renal data were found to have significantly worse survival, identification of outlying individuals through robust joint modelling may aid nephrologists to improve patient's survival. A simulation study was also undertaken to explore the difference between robust and standard joint models in the presence of increasing proportions and extremity of longitudinal outliers. More efficient and reliable estimates were obtained by robust joint models with increasing contrast between the robust and standard joint models when a greater proportion of more extreme outliers are present. Through illustration of the gains in efficiency and reliability of parameters when outliers exist, the potential of robust joint modelling is evident. The research presented in this thesis highlights the benefits and stresses the need to utilise a more robust approach to joint modelling in the presence of longitudinal outliers.

Application of GPCR Structures for Modelling of Free Fatty Acid Receptors

Five G protein-coupled receptors (GPCRs) have been identified to be activated by free fatty acids (FFA). Among them, FFA1 (GPR40) and FFA4 (GPR120) bind long-chain fatty acids, FFA2 (GPR43) and FFA3 (GPR41) bind short-chain fatty acids and GPR84 binds medium-chain fatty acids. Free fatty acid receptors have now emerged as potential targets for the treatment of diabetes, obesity and immune diseases. The recent progress in crystallography of GPCRs has now enabled the elucidation of the structure of FFA1 and provided reliable templates for homology modelling of other FFA receptors. Analysis of the crystal structure and improved homology models, along with mutagenesis data and structure activity, highlighted an unusual arginine charge pairing interaction in FFA1-3 for receptor modulation, distinct structural features for ligand binding to FFA1 and FFA4 and an arginine of the second extracellular loop as a possible anchoring point for FFA at GPR84. Structural data will be helpful for searching novel small molecule modulators at the FFA receptors.

Analytic modelling of the galactic fountain in the Milky Way potential

Le galassie a spirale, come la Via Lattea, sono caratterizzate dalla presenza di gas freddo e formazione stellare e vengono perciò chiamate star-forming. Per creare nuove stelle è necessaria una sufficiente riserva di gas, la cui disponibilità governa l’evoluzione della galassia stessa. Finora, non è stato individuato con certezza un meccanismo che possa alimentare la formazione di nuove stelle nelle galassie star-forming. Una delle possibili sorgenti di tale gas è l’alone galattico caldo (corona galattica) il cui raffreddamento e successivo accrescimento possono essere stimolati dal processo di fontana galattica. L’esplosione di supernovae porta nubi di gas freddo in orbita al di sopra del disco stellare; queste nubi raggiungono altezze dell’ordine del kiloparsec, interagendo con la corona di gas caldo. Il moto delle nubi all’interno di un mezzo meno denso comporta l’instaurarsi dell’instabilità di Kelvin-Helmholtz, che ’strappa’ gas dalle nubi e causa la condensazione di materia coronale. Quest’ultima viene quindi accresciuta e, ricadendo sul disco, trasferisce nuovo materiale alla galassia e ne alimenta la formazione stellare. Lo scopo di questa tesi è derivare un modello analitico di fontana galattica che consenta di ottenere una formulazione analitica per il tempo orbitale, cioè il tempo richiesto alle nubi per ricadere sul disco galattico. Infatti, più u tempo le nubi impiegano per attraversare il materiale coronale caldo e ricadere sul disco, più materiale viene accresciuto durante l’orbita. Conoscendo i tempi orbitali sarebbe possibile calcolare il tasso di accrescimento legato al fenomeno di fontana e studiarne l’andamento con il raggio del disco. Questo modello potrebbe rivelarsi utile per lo studio dell’impatto della fontana nell’evoluzione globale del disco galattico e della chimica dell’intera galassia.

Air Temperature Distribution and Energy-balance Modelling of a Debris-covered Glacier

Near-surface air temperature is an important determinant of the surface energy balance of glaciers and is often represented by a constant linear temperature gradients (TGs) in models. Spatiotemporal variability in 2 m air temperature was measured across the debris-covered Miage Glacier, Italy, over an 89 d period during the 2014 ablation season using a network of 19 stations. Air temperature was found to be strongly dependent upon elevation for most stations, even under varying meteorological conditions and at different times of day, and its spatial variability was well explained by a locally derived mean linear TG (MG–TG) of −0.0088°C m−1. However, local temperature depressions occurred over areas of very thin or patchy debris cover. The MG–TG, together with other air TGs, extrapolated from both on- and off-glacier sites, were applied in a distributed energy-balance model. Compared with piecewise air temperature extrapolation from all on-glacier stations, modelled ablation, using the MG–TG, increased by <1%, increasing to >4% using the environmental ‘lapse rate’. Ice melt under thick debris was relatively insensitive to air temperature, while the effects of different temperature extrapolation methods were strongest at high elevation sites of thin and patchy debris cover.

On the modelling of fire atmosphere/sprinkler spray interacting using an Euler-Lagrange framework

An Euler-Lagrange particle tracking model, developed for simulating fire atmosphere/sprinkler spray interactions, is described. Full details of the model along with the approximations made and restrictions applying are presented. Errors commonly found in previous formulations of the source terms used in this two-phase approach are described and corrected. In order to demonstrate the capabilities of the model it is applied to the simulation of a fire in a long corridor containing a sprinkler. The simulation presented is three-dimensional and transient and considers mass, momentum and energy transfer between the gaseous atmosphere and injected liquid droplets.

CFD modelling of magnetic levitation casting

Abstract not available

Multi-physics modelling of materials processes on high performance parallel computers

Abstract not available