OBJECTIVE

2

To evaluate the quality of Zhachong-13 pills.

METHODS

2

The contents of agarotetrol， costunolide， hypaconitine， dehydrocostus lactone， mesaconitine， aconitine， liquiritigenin， glycyrrhizic acid， ellagic acid， liquiritin， corilagin， and gallic acid in 18 batches of samples were determined by high-performance liquid chromatography-tandem triple quadrupole mass spectrometry （HPLC-MS/MS） method. Cluster analysis （CA） was performed using the Microbioinformatics online analysis platform. Principal component analysis （PCA） and orthogonal partial least squares discriminant analysis （OPLS-DA） were performed using SIMCA 14.1 software. Random forest， extremely randomized trees， and extreme gradient boosting machine learning models were constructed using the SPSSAU online analysis platform. The comprehensive quality of the samples was evaluated by the entropy weight-technique for order preference by similarity to ideal solution （TOPSIS） method.

RESULTS

2

The contents of the above 12 components were 104.331-694.382， 652.217-1 085.075， 6.051-130.980， 612.174-1 417.179， 1.018-32.239， 0.953-24.967， 65.552-652.239， 2 019.412-5 385.807， 52.867-4 990.450， 156.307-883.180， 10.142-475.722， and 1 102.863-4 826.396 μg/g， respectively. The results of CA and PCA showed that the 18 batches of samples were divided into two categories： S1-S6， S10-S15 were grouped into one category， while S7-S9 and S16-S18 were grouped into another category. The OPLS-DA results showed that the variable importance in projection values of mesaconitine， glycyrrhizic acid， liquiritin， ellagic acid， hypaconitine， liquiritigenin， agarotetrol， gallic acid， and dehydrocostus lactone were greater than 1. The results of the three machine learning models showed that ellagic acid and hypaconitine had relatively large differential contributions. The entropy weight-TOPSIS results showed that the overall quality of samples from manufacturer B was superior to that of the other four manufacturers.

CONCLUSIONS

2

The established method is stable and feasible， and can be used to evaluate the quality of Zhachong-13 pills， in combination with multivariate statistical analysis. Ellagic acid and hypaconitine are differential markers that affect its quality.