Integrative analysis of time course metabolic data and biomarker discovery

摘要

Over the past few years, there has been a significant development in high-throughput omics technologies e.g. metabolomics, transcriptomics, genomics, epigenomics and proteomics along with a growing interest into joint modeling of multi-omic data [ , ]. In metabolomics, several approaches are used to understand the response of a biological system as a function of an internal or external perturbation by monitoring “the chemical fingerprints that specific cellular processes leave behind" [ ]. These chemical fingerprints are most commonly interrogated in terms of metabolite (i.e. low weight molecules) concentration, structure and transformation pathways (i.e set of chemical reactions) in order to identify biomarkers related to the studied process. Biomarker discovery consists of identifying a metabolite that has significant association patterns with a particular phenotype (disease, clinical variables, physical trait, etc) and that can be thus used as an indicator of that specific phenotype. Typical experimental platforms use analytical techniques such as nuclear magnetic resonance spectroscopy (NMR) [ ] and mass spectrometry (MS) [ ] to generate appropriate spectral metabolomic profiles of the studied biological system.