中文摘要：

旨在建立组分中药组方筛选新方法，以期客观评估多组分内在关系及优化配比。以流感病毒神经氨酸酶（NA）活性抑制率为指标，采用Box．Behnken响应曲面设计法，对双黄连注射液（SHL）中主要指标成分进行抑制NA活性筛选，探讨其活性组分间的相互作用关系，并以SHL为参照，预测组合成分的最佳配比。结果表明，原方中绿原酸、隐绿原酸、咖啡酸和黄芩苷有较强的NA活性抑制作用，且绿原酸与隐绿原酸间存在拮抗作用，咖啡酸对其他组分具有协同增效作用；连翘酯苷A、连翘苷等未见明显（或协同）NA活性抑制作用；拟合的模型预测组分最佳配比为绿原酸一隐绿原酸一咖啡酸一黄芩苷（107gg·mL-1：279tμg-mL-1．99gg·mL-1：92gg·mL-1，经实验验证，其NA活性抑制作用强于SHL（P〈0．05）。Box．Behnken响应曲面设计法具有普筛高效、预测准确等优势，适合用于组分中药组方的筛选与优化配比研究，为有效推进现代组分中药的发展提供了技术支持。

英文摘要：

This study aimed to establish a novel method to screen out the combined components of multifractions traditional Chinese medicine （TCM）, so that the internal relationship between multi-ingredients could be objectively assessed and the proportioning ratio could be optimized. Taking antiviral effect on neuraminidase activity of influenza virus as the evaluating indicator and using Box-Behnken response surface methodology, the main effective ingredients of Shuanghuanglian injection （SHL） were screened. Meanwhile, the relationship between active ingredients was discussed. Taking SHL as a comparison, the optimum proportioning ratio was predicted. The results indicated that chlorogenic acid, cryptochlorogenic acid, caffeic acid and baicalinhave comparatively strong antiviral activity against influenza virus. Moreover, antagonistic action existed between chlorogenic acid and cryptochlorogenic acid, whereas synergistic action between caffeic acid and other components. The optimum proportioning ratio resulted from fitted model is： chlorogenic acid, cryptochlorogenic acid, caffeic acid and baicalin （107/xg.mL-1 ： 279 μg.mL-1 ： 7.99 Ixg.mL-1 ： 92 μg.mL-1）. The antiviral activity of the recombined components is stronger than that of SHL, which was consistent with the experiment results （P 〈 0.05）. Box-Behnken response surface methodology has the advantages of general-screening, high-performance and accurate-prediction etc, which is appropriate for screening the combined components of multi-fractions TCM and the optimization of the proportioning ratio. The proposed method can serve as a technological support for the development of modem multi-fractions TCM.