Objective: To explore the application and rationality of the nonlinear mixed effect model (NONMEM) in quantitative analysis for finding an optimal combination based on the public published data. Method: Based on the original data of Danshou tang formulation in the published literature

an orthogonal design

with 1-level (used) and 2-level (not used) in L12

was used to assign experimental groups of the combinations of components

the embryos absorption rate was the primary endpoint

mathematical model was build up with the nonlinear mixed effect model (NONMEM) for the experiment. The importance and effectiveness of each component were analyzed in combinations and the bias was evaluated by 4 charts. Result: The whole formulation of Danshou tang showed addition action

the contribution to combined effect of each component in order was A

E

B (P<0.05). The best combination would be the ABCEF theoretically

and 4 diagrams revealed the reliability and the stability of the model. Conclusion: More information about prescription compatibility can be obtained by the orthogonal simulation method with NONMEM.