Characterization and Discrimination of Volatile Compounds in Honeysuckle From Different Plots Using E-Tongue, E-Nose, GC-IMS, and HS-SPME-GC-MS Combined With Random Forest Algorithm.

Honeysuckle samples from various plots in Pingyi County, Linyi City were selected and analyzed by integrating electronic tongue (E-tongue), electronic nose (E-nose), HS-SPME-GC-MS, and GC-IMS technologies with multivariate statistical analysis methods, a systematic analysis of the flavor characteristics and plot differentiation patterns of volatile organic compounds (VOCs) was conducted in the study. The results indicated that the E-tongue provided distinct sensor response patterns that correlated with taste attributes such as sweetness and bitterness, effectively discriminating samples from different plots. E-nose analysis further showed that aroma-related response intensities exhibited significant plot-based variation. Through HS-SPME-GC-MS and GC-IMS techniques, a total of 203 volatile components were identified. Based on HS-SPME-GC-MS data, using VIP > 1 and OAV ≥ 1 as screening criteria, seven key volatile markers were determined. From the GC-IMS data, 24 characteristic compounds were screened out using the criterion of VIP > 1, with aldehydes comprising the highest proportion. Furthermore, the random forest (RF) algorithm was employed to integrate these two sets of data. Monte Carlo cross-validation showed that the model accuracy reached 1.0, successfully identifying 15 key volatile compounds. This study provided a theoretical foundation and data support for optimizing the quality evaluation system of honeysuckle and innovating origin traceability technology through multi-dimensional flavor analysis and key marker screening.