Search

Severe thermal injury results in a profound hypermetabolic response and is associated with increased morbidity, mortality and delayed rehabilitation of burn survivors. Serum 1H-NMR metabolomics was used to examine early global metabolic changes in the response to severe thermal injury (>15% TBSA) in young adults (16-64 years) and older adults (< 15% TBSA). Early changes in the metabolome reflected hypoxic metabolism, hyperglycaemia, increased ketogenesis, peripheral lipolysis and increased energy production in both cohorts. Early metabolic profiles from the young adult group were used to construct OPLSDA models that could discriminate with high accuracy between outcome groups. Models from 0-24hrs serum samples predicted survival (AUC 0.92), whilst models from 24-96hrs samples predicted Multiple organ failure (MOF) (AUC 0.92) and sepsis (AUC 0.89). Untargeted LC-MS metabolomics was applied to study the longitudinal changes in the serum metabolome after severe thermal injury in 13 young adults, from admission until 6-months post-injury. Univariate ANOVA analysis revealed significant changes in 432 metabolite features, affecting 35 distinct classes, representing global metabolic disturbance. Changes in 300 lipid metabolite features may represent a ‘lipid storm’ in serum after severe thermal injury. Novel areas of metabolism and metabolites were identified as putative biomarkers warranting further targeted study.

We introduce Wilson's, or Polchinski's, exact renormalization group, and review the Local Potential Approximation as applied to scalar field theory. Focusing on the Polchinski flow equation, standard methods are investigated, and by choosing restrictions to some sub-manifold of coupling constant space we arrive at a very promising variational approximation method. Within the Local Potential Approximation, we construct a function, C, of the coupling constants; it has the property that (for unitary theories) it decreases monotonically along flows and is stationary only at fixed points - where it 'counts degrees of freedom', i.e. is extensive, counting one for each Gaussian scalar. In the latter part of the thesis, the Local Potential Approximation is used to derive a non-trivial Polchinski flow equation to include Fermi fields. Our flow equation does not support chirally invariant solutions and does not reproduce the features associated with the corresponding invariant theories. We solve both for a finite number of components, N, and within the large N limit. The Legendre flow equation provides a comparison with exact results in the large N limit. In this limit, it is solved to yield both chirally invariant and non-invariant solutions.