For the analysis of complex samples, comprehensive two-dimensional gas chromatography coupled with time-of-flight mass spectrometry (GC x GC - TOFMS) is a powerful instrumental platform. Chemometric approaches play a pivotal role in the analytical workflow for the translation of the raw data into useful information. We are exploring and developing advanced approaches for non-targeted discovery-based analysis of cross-sample comparisons of GC x GC - TOFMS data, coupled with robust deconvolution, identification and quantification of meaningful analytes. Recent advances in tile-based Fisher-ratio (F-ratio) analysis will be presented, a statistically-based data mining technique to discover analytes that distinguish sample classes based upon the experimental design. The tile-based F-ratio software substantially improves chemical selectivity in the discovery process for the determination of an analyte hit list. Most recently, we have applied combinatorial theory via null distribution analysis to determine a statistically robust F-ratio threshold that confidently guides the mining of the hit list. Important features discovered by F-ratio analysis are further analyzed using complementary chemometric deconvolution methods. This analytical platform combines experimental design, instrumental and computational methods for rapid and high-quality extraction of the most useful information from complex discovery-based GC x GC - TOFMS studies. Specifically, this analytical platform will be showcased in a biomarker discovery application to a previously validated yeast metabolomics experiment, in which the tile-based F-ratio software reduced 70 GC x GC - TOFMS chromatograms to 96 quantified metabolites which differentiate the yeast metabolomic states, concurrent with essentially no false positives.

Discovery-based analysis of GC x GC - TOFMS data using tile-based Fisher ratio software and combinatorial threshold determination

251st National Meeting (2016)

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.