Monti and colleagues analyzed flower from 18 cannabis varieties grown under standardized conditions. Their targeted high-resolution mass spectrometry workflow measured a panel of 15 cannabinoids. A parallel untargeted workflow detected 19 additional compounds and used computational tools to help assign structures. This produced a richer fingerprint for comparing the study samples.
An instrument can reproducibly measure a feature at a given mass and retention time even when the precise structure remains tentative. Possible isomers can share a molecular formula. A proposed structure carries more weight when fragmentation and retention behavior agree with a suitable authentic reference. The authors discussed the tentative status of assignments made through the untargeted workflow.
The study used principal component analysis to visualize patterns, including subgroups beyond the usual THC-to-CBD classification. A plot of these samples describes their similarities under the study conditions. An inference about an unknown sample’s source would need a reference population, independent validation and an estimate of how often unrelated samples cluster together. A laboratory report should identify which findings are quantified compounds and which are tentative features.