Sample-Specific Metabolites Library with Retention Neighbor: an Improved Identification and Quantitation Strategy for Gas Chromatography–Mass Spectrometry-Based Metabolomics

In this study, the idea of sample-specific metabolites library with retention neighbor was presented for better identification and quantitation of metabolites. Compared with universal libraries, the established library was found to have significantly improved match indices (similarity, reverse, and probability), indicating better identification of metabolites. For quantitative metabolic profiling of temperature treated tobacco samples, the established method integrated and aligned 175 metabolites with no missing value, better than the widely used metabolomics software including XCMS Online, MetaboliteDetector, SpectConnect, and ChromaTOF. Partial least squares discriminant analysis models were also used for quantitative evaluation and the established method produced better values of model parameters than the compared software, indicating a better model was established using the present method.