毛橘红总黄酮指纹图谱与其抗氧化活性的谱效关系研究

Research of Spectral Efficiency Relationship Between Total Flavonoids Fingerprints and Antioxidant Activity of Exocarpium Citri Grandis

【目的】研究毛橘红总黄酮指纹图谱与其抗氧化活性的谱效关系。【方法】采用高效液相色谱(HPLC)法制备毛橘红总黄酮指纹图谱,按中国药典法进行抗氧化作用强度的测定。应用灰关联(GM)数学模型计算各色谱峰与抗氧化活性的灰关联度,采用逐步多元回归对指纹色谱峰进行回归、筛选,并结合广义回归人工神经网络(GRNN)进行抗氧化性预测。【结果】筛选出对抗氧化作用有统计学意义(P<0.05)的4个峰,GRNN预测模型显示筛选后的预测误差显著小于筛选前的预测误差(P<0.05)。【结论】灰关联与逐步多元回归结合的方法可阐明毛橘红总黄酮指纹图谱与抗氧化活性间的谱效关系,GRNN法可较好地预测毛橘红谱峰的抗氧化强度。

Objective To investigate the spectral efficiency relationship between total flavonoids fingerprints and antioxidant activity of Exocarpium Citri Grandis. Methods The total flavonoids fingerprints of Exocarpium Citri Grandis were prepared by using high pressure liquid chromatography （HPLC） method. We determined the antioxidant activity of Exocarpium Citri Grandis according to the Chinese Pharmacopoeia method, and caiculated the grey relation grade between different spectrum peaks and their antioxidant activity with the Grey-relation Model （GM）. The stepwise multiple regression method was used for regression analysis and screening of fingerprint peaks, and the General Regression Neural Network （GRNN） was used to predict the antioxidant activity. Results Four peaks which had statistically significant difference of the antioxidant effect （P 〈 0. 05 ） had been screened out, and the GRNN prediction model displayed that the prediction deviation after the screening was significantly less than that before using the model （P 〈 0. 05 ）. Conclusion The combination of the grey relation analysis and stepwise multiple regression analysis can be used to clarify the spectral efficiency relationship between total flavonoids fingerprints and antioxidant activity of Exocarpium Citri Grandis, and by adopting the GRNN model we can predict the antioxidant efficiency of the peaks.