Bioinformatic approaches to identify genomic, proteomic and metabolomic biomarkers for the metabolic syndrome

Advances in technology have turned modern biology into a data-intensive enterprise. The advent of high-output technologies like Microarrays and Next-generation sequencing technologies has resulted in researchers grappling not just with huge volumes but also multiple types of data. While generation and storage of high-quality data are an important research focus, it is increasingly recognized that translating data into actionable information and insight is a critical research challenge. To infer reliable conclusions from the data, it is often necessary to integrate large amounts of heterogeneous data with different formats and semantics. Given the breadth and volume of data involved, this goal is best achieved through automated methods and tools for data integration and workflow management. This thesis presents automated strategies that combine bioinformatics and statistical methods to identify novel biomarkers in high-throughput OMICs datasets pertaining to the metabolic syndrome and to gain mechanistic insight into the underlying biological processes. An underlying theme in this thesis is data-driven approaches that generate plausible hypothesis which is followed by experimental verification