Use this URL to cite or link to this record in EThOS: http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.669506
Title: Novel statistical and bioinformatic tools for identifying predictive metabolic biomarkers in molecular epidemiology studies
Author: Posma, Joram Matthias
ISNI:       0000 0004 5369 0439
Awarding Body: Imperial College London
Current Institution: Imperial College London
Date of Award: 2014
Availability of Full Text:
Access from EThOS:
Full text unavailable from EThOS. Please try the link below.
Access from Institution:
Abstract:
A top-down systems biology approach investigating metabolic responses to external stimuli or physiological processes requires multivariate statistical tools to identify metabolites associated with the global biochemical changes in a supra-organism. In this thesis I describe several tools I have developed to improve or supplement currently used methods in molecular epidemiology studies. First, I describe the MetaboNetworks toolbox which is able to create custom, multi-compartmental metabolic reaction networks for a supra-organism, combining both mammalian and microbial reactions. These networks are essentially a summary of the supra-organisms homeostatic signature. Second, I describe a novel statistical spectroscopy approach called STORM which aids in the elucidation of unknown biomarker signals in 1H NMR spectra. Third, I describe the Metabolome-Wide Association Study on obesity in U.S. and U.K. populations. Many novel metabolic associations with obesity are described in a systems framework, among which metabolites associated with energy, skeletal muscle, lipid, amino acid and gut microbial metabolism. Last, I describe a new multivariate approach to adjust for confounders, CA-OPLS. Correcting for confounders is an essential aspect in molecular epidemiology studies as metabolites can be related to a variety of factors such as lifestyle, diet and environmental exposures which or may not be causally related to disease risk. In developing CA-OPLS another aim was to simultaneously eliminate/minimize the effects of different types of sampling bias which are often not taken into account in modelling metabonomics data with current methods.
Supervisor: Nicholson, Jeremy; Elliott, Paul Sponsor: Medical Research Council ; Public Health England
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID: uk.bl.ethos.669506  DOI: Not available
Share: