Datamining metabolomics: the convergence point of non-target approach and statistical investigation

Metabolomics studies are governed by the latest technological developments in analytical chemistry, statistics and bioinformatics. Ion Cyclotron Resonance Fourier Transform Mass Spectrometry (ICR-FT/MS) at high magnetic field is a new generation of mass spectrometers, with ultra-high resolution and mass accuracy enabling molecular resolution on the isomeric space. This thesis presents strategic approaches to analyze large datasets as obtained with ultra-high resolution instruments. The approach foresees two fundamental steps: the first one consisting in the data reduction followed by data analysis and visualization and the second one to confront the real experimental mass values with putative structures and metabolite profiles. After setting the basis for this approach the strategy was used for various issues: environmental-biogeography, food chemistry and biomedical/diseases diagnosis. Even though the approach presents limitations, it provides high potential in analyzing and extrapolating information from high complex datasets and was applied for different applications to evaluate its universal structural characteristic. One solution to these limitations is the combination of different analytical techniques (separation/spectrometry/spectroscopy) which generates such a bright view of the data from a sample that evaluation is challenging to a statistician.