Résumé
The paper reports the development of a quality evaluation method for Angelica different processed products. The data of high-performance liquid chromatography, water, total ash and extract were analyzed with SPSS Clementine 11.0 software. Discriminant analysis (DA) established the classification model and parameter for Angelica different processed products. Fish's discriminant functions of Angelica different processed products were generated using 8 predictor variables selected from 59 indexes. The correct rate of discriminating back substitution is 96.7%. Angelica different processed products can be accurately and reliably recognized and validated with DA of SPSS Clementine 11.0 software.
Résumé
<p><b>OBJECTIVE</b>To establish a chemical fingerprint method for reorganizing and validating angelica different processed products.</p><p><b>METHOD</b>A high-performance liquid chromatographic method was developed to establish the fingerprint. Principal component analysis, hierarchical cluster analysis and discriminate analysis were applied to study HPLC finger printing and chemical pattern reorganization.</p><p><b>RESULT</b>There were difference of characteristic peaks and its relative peak area of HPLC fingerprints between different processed products. Fish's discriminate functions were generated by using six selected predictor variables, the tested samples of different processed products were classified with 100% accuracy, and discriminate analysis plots for the five groups were well-resolved.</p><p><b>CONCLUSION</b>The developed HPLC finger print, combined with chemometrics, can accurately identify and validate angelica different processed products, the research provide theoretical basis for the processing mechanism and quality assess of angelica different processed products.</p>