Decoding honey-sweet flavored flue-cured tobacco from Guizhou with data science and flavoromics by volatile and cell wall components.

Abstract

Nicotiana tabacum L. is often called tobacco. The aroma of flue-cured tobacco (FCT) varies according to the origin and grade. In this study, volatiles and plant cell wall components (CWC) were used to differentiate aroma types and grades of FCT, with a focus on the honey-sweet flavored FCT from Guizhou, China. Volatiles were analyzed by headspace solid-phase microextraction gas chromatography/mass spectrometry, while CWC (cellulose, hemicellulose, pectin, lignin) were quantified. Results indicated that upper leaf (Grade B) tobacco contained higher volatile levels than middle leaf (Grade C). Multivariate analyses-Principal component analysis (PCA), orthogonal partial least squares discriminant analysis (OPLS-DA), and logistic regression (LR) identified 27 key volatiles contributing to aroma differentiation in FCT origin. By combining the screened volatiles with odor activity value, the most important key aroma compounds that distinguish Guizhou honey-sweet flavored from other origins were β-cyclocitral and 1-nonanal. The CWC showed significant variation across origins or grades. Machine learning models (e.g., LR with 96.5% accuracy) effectively distinguished the origin of FCT. This study pioneers the integration of machine learning with molecular sensory science to decode the unique honey-sweet flavor of Guizhou flue-cured tobacco, addressing a critical gap in linking volatile biomarkers to regional terroir. This methodology provides a way to evaluate tobacco quality and aroma characteristics.