- Development and optimization of 3 high resolution non-targeted LC-MS methods.
- Metabolic fingerprints of cocoa shell and cocoa nibs from 63 cocoa been samples.
- Defining selection criteria for the potential key metabolites.
- Identification of 18 potent key metabolites for cocoa shell detection.
- SPLS regression is used for the prediction of the cocoa shell content.
The determination of cocoa shell content (
L.) in cocoa products using a metabolomics approach was accomplished via high performance liquid chromatography quadrupole time-of-flight mass spectrometry (HPLC-QTOF-MS). The developed method was used to separately analyze the polar and non-polar metabolome of the cocoa testa (cocoa shell) and the cocoa cotyledons (cocoa nibs) of cocoa samples from 15 different geographic origins, harvest years, and varieties in positive and negative ion mode. Potential key metabolites were selected which are exclusively contained in the cocoa shell or with significant higher concentration in the cocoa shell than in the cocoa nibs. The pool of potential key metabolites was filtered by established selection criteria, such as temperature stability, fermentations stability, and independence from the geographic origin. Based on these key metabolites an inverse sparse partial least square regression (SPLS) was used for the prediction of the cocoa shell content.