Chemical profiling and clustering of various dried cannabis flowers revealed by volatilomics and chemometric processing

Abstract Cannabis flower scent is one of the key characteristics of the cannabis plant. The diverse scents impact user experiences and offer medicinal benefits. These scents originate from volatile compounds, particularly terpenes and terpenoids. This study characterized the volatile profile of 19 different dried cannabis flowers using gas chromatography-mass spectrometry coupled with headspace-solid phase microextraction (HS-SPME-GC-MS). A total of 75 compounds were identified, including alcohols, aldehydes, benzenes, esters, ketone, monoterpenes, monoterpenoids, sesquiterpenes, and sesquiterpenoids. Cluster analysis was able to group the 19 cannabis cultivars into five clusters based on volatile chemotypes using chemometric techniques of hierarchical cluster analysis (HCA) and principal component analysis (PCA). Potential discriminant markers of each cultivar were then analyzed using a supervised partial least squares discriminant analysis (PLS-DA) verified through Variable Importance in Projection values (VIP), identifying twenty discriminant markers. In addition, the correlations among 75 volatile compounds were also obtained. The findings of this study provide a valuable database of single cannabis cultivars, useful for identifying individual strains and verifying their quality. Clustering the cultivars by volatile chemotype can be used for the classification of cannabis in the market. The results of this study are expected to be a starting point for further cannabis breeding programs to expand knowledge of this plant. Furthermore, the proposed method is applicable to other aroma plants in the future.