A Comparison Studies On The Design Of Drug Combination Experiment

As the science of isolation technology and chemical synthetic capability advance,antiviral drug combinations are increasingly used to reduce possible drug-resistant viral mutant and reduce cytotoxicity. However, because biological systems as well as dose-effect models are exceedingly complex, the combined antiviral drug effect is generally hard to assess. One important reason is due to the complex interactions between biological systems and drug molecules. To select the optimal combination,preclinical trials in vitro are usually conducted.Good experiments designs frequent-ly can reduce the experiments’ time and cost, as the same time, provide enough information for subsequent data analysis.Owing to the complex interactions of drug combinations, lots of statisticians proposem many effective analysis models.Different designs are put forward according to these models. However, owing to cost and ef-ficiency considerations, both the number of runs and the range of drug dosages in the experiments are limited. Therefore, the application of computer simulation technology is particularly important.Especially two drugs combinations experimen-t design.But when increasing the drug kinds in drug combination experiment, the relevant method and theory become more and more difficult the existing methods are a few.This article aims at Hill-based-polynomial model which is widely used in drug experiment, proposes mixture-dose combine experiment design. The drug combina-tion experiment is divided into two dimensions:1)When the total dose fixed, various kinds of drugs proportions of mixed ex-periment in drug combination,can use mixture design.2)When fix the drug proportions, the influence of different dose of drug combi-nation of response expresses as a quadratic curve, can use the optimal design.At the same trail scenario,using computer simulation method,combining the random search algorithm to solve the parameter optimal value.By computation the MSE and residuals in each design scenario, draw a conclusion that the mixture-dose combination design is good in some ways from the fitting results.