OBJECTIVE  To establish a method for the classification of Lonicerae Japonica Flos and Lonicerae Flos, and the identification of Lonicerae Japonica Flos whether from authentic producing areas.

METHODS  The data from nuclear magnetic resonance(NMR) was integrated with the one-stop analysis software ChemPattern 2020 to pattern recognition of Lonicerae Japonicae Flos, and the chlorogenic acid of the primary active component in Lonicerae Japonicae Flos and Lonicerae Flos was quantified.

RESULTS  The fingerprints of 60 batches of Lonicerae Japonicae Flos and Lonicerae Flos were established. The results of the principal component analysis indicated that the cumulative variance contribution rate of the top 3 principal components was 96.97%. The 60 batches of samples were categorized into 2 distinct groups, revealing a notable difference between Lonicerae Japonicae Flos and Lonicerae Flos. Furthermore, the orthogonal partial least squares-discriminant analysis demonstrated test parameters approaching 1, and 60 batches of samples were clustered into 3 groups. This analysis confirmed the identification of Lonicerae Japonica Flos and Lonicerae Flos, and the Lonicerae Japonica Flos whether from authentic producing areas. Additionally, a recognition model based on support vector machine(SVM) was developed. The kernel function employed was radial basis function, utilizing a 10-fold cross-validation approach, under conditions of uniform data preprocessing, this established SVM model achieved an impressive recognition rate of 100%. Meanwhile, chlorogenic acid levels in samples from thirteen different regions were determined using ¹H-NMR spectroscopy.

CONCLUSION  ¹H-NMR data combined with pattern recognition effectively distinguish Lonicerae Japonica Flos and Lonicerae Flos and Lonicerae Japonica Flos whether from authentic producing areas.