应用支持向量机筛选香丹注射液的抗缺氧有效部位

Simulation Study of Anti-hypoxia Active Fraction of Xiangdan Injection by Support Vector Machine

目的:探讨香丹注射液的抗缺氧有效部位。方法:采用小鼠常压耐缺氧实验,基于正交设计,建立复方药效的支持向量机预测模型;并经实验验证,确定香丹注射液的抗缺氧有效部位。结果:本研究所建立的复方预测模型可以对复方不同组合进行药效预测。结论:通过模型预测和实验验证,多个配伍组的药效优于原方,尤以ADE效果最佳。

Objective： To investigate the anti-hypoxia active fraction of Xiangdan injection. Methods： The anti-hypoxia experiment under normal pressure with mice was adopted on the basis of orthogonal design, and a support vector machine model was established for pharmacodynamics prediction of TCM prescriptions. Moreover, the anti-hypoxia active fraction of Xiangdan injection was determined by experimental confirmation. Results ： The prediction model established in this study could predict the drug actions of different formulas. Conclusion ： As indicated in model prediction and experimental confirmation, the phannacodynamic actions of several formulas are superior to that of the original formula, and ADE has the best effect.