Discriminant analysis of raw and honey baked Farfarae Flos from different regions by HPLC fingerprint analysis combined with chemical pattern recognition / 中草药

ABSTRACT

Objective To establish an HPLC fingerprint of raw and honey baked Farfarae Flos for its quality control and samples differentiation. Methods An HPLC method has been developed for the fingerprinting and evaluation of 36 batches of raw and honey baked Farfarae Flos collected from different locations. The Similarity Evaluation System for Chromatographic Fingerprint of TCM (2012A edition) was used to evaluate the similarity of 36 batches. The difference between raw and honey baked Farfarae Flos was identified by chemical pattern recognition methods including hierarchical cluster analysis (HCA), principal component analysis (PCA), and partial least squares discriminate analysis (PLS-DA). Results A standard fingerprint containing 20 common peaks was constructed from 36 batches of raw and honey baked Farfarae Flos, and identified 10 of them. The similarity of all batches with reference fingerprint was between 0.723-0.984. The similarity of 16 batches of raw Farfarae Flos was between 0.862-0.998, and the similarity of 20 batches of honey baked Farfarae Flos was between 0.687-0.993. HCA, PCA and PLS-DA results demonstrated that there were obvious distinction between raw and honey baked Farfarae Flos. According to the VIP plot, ten constituents including gallic acid, chlorogenic acid, isochlorogenic acid A, and tussilagone were primarily response for the discrimination. Conclusion The combination of HPLC fingerprint and chemical pattern recognition could provide a comprehensive reference for the quality control and quality evaluation of raw and honey baked Farfarae Flos.