The Dark side of Obesity: Multi-omics analysis of the dysmetabolic morbidities spectrum
Abstract: Obesity is one of the most prevalent clinical conditions worldwide and is associated with a wide spectrum of dysmetabolic comorbidities. Complex cardio-metabolic disease cohorts, such as obesity cohorts are characterised by population heterogeneity, multiple underlying diseases status and different comorbidities’ treatment regiments. The systematic collection of multiple types of clinical and biological data from such cohorts and the data-analysis in an integrative manner is a challenging task due to the variables’ dimensionality and the lack of standardised know-how of post-processing. The main resource of this thesis has been the BARIA cohort, a detailed collection over time of multiple omics and demographic data from participants in bariatric surgery. BARIA datasets included plasma metabolites, RNA from hepatic, jejunal, mesenteric and subcutaneous adipose tissues and gut microbial metagenome, besides biometric data. The work presented in this thesis included the development of a systems biology integrative framework based on BARIA that (i) utilised unsupervised machine learning algorithms, self-organizing maps in particular, and multi-omics integrative frameworks, the DIABLO library, in order to stratify the BARIA heterogeneous obesity cohort and predict the bariatric surgery’s outcome. The thesis covered how BARIA can be the onset for (ii) studying molecular mechanisms related to type 2 diabetes (T2D) and G-protein coupled receptors (GPCRs) and for identifying a minimal set of biomarkers for obesity’s comorbidities such as (iii) non-alcoholic fatty liver disease (NAFL) and (iv) gallstones formation after bariatric surgery. The results indicated that the metabotypes comprising a bariatric surgery cohort exhibited a concrete metabolic status and different responses over time after the bariatric surgery. It has been demonstrated how obesity and T2D associated metabolites, such as 3-hydroxydecanoate, can increase inflammatory responses via GPCRs molecular activation and signalling. Last but not least, minimal sets of both evasive and non-evasive multi-omic discriminatory biomarkers for obesity’s dysmetabolic morbidities (NAFLD and gallstones after bariatric surgery) were obtained. Taking into consideration all the findings, this thesis presented how data-driven approaches can be used for studying in-depth heterogeneous cohorts, hereby facilitating early diagnosis and enabling potential preventive actions.
This dissertation MIGHT be available in PDF-format. Check this page to see if it is available for download.