Food authentication in the routine laboratory: Determination of the geographical origin of white asparagus using a simple targeted LC-ESI-QqQ-MS/MS approach

In the last decade, research efforts to develop non-targeted metabolomics strategies, especially for the detection of various food authenticity parameters, have increased significantly. However, such methods are hardly implemented in routine analysis so far, which is due to the relatively expensive equipment as well as the lack of reference databases. The present study describes the downstream process for converting a non-targeted metabolomics study into a simple targeted approach suitable for routine analysis. Using 20 marker compounds and random forest algorithm (RF), the geographical origin of 140 asparagus samples from five regions (Germany, Poland, the Netherlands, Greece/Spain, and Peru) was determined. With a prediction accuracy of 76%, the targeted method provided only a slightly worse result than the non-targeted method and significantly better results than the isotope ratio mass spectrometry (IRMS), which is currently used for comparable issues.