[Simulation study of anti-hypoxia active fraction of Xiangdan injection by support vector machine].

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 pharmacodynamic actions of several formulas are superior to that of the original formula, and ADE has the best effect.