论文部分内容阅读
介绍了混杂样本剖析的Gram-Charlier级数模型法,并应用其实现药物代谢多态性的合理分型。给出两样本混杂的G-C级数模型,用具有自定义非线性模型拟合功能的计算机软件,借助拟牛顿法提出分步完成模型参数的非线性最小二乘估计,通过理论频数与实际频数间的拟合优度检验判定模型的合理性,用秩和检验推断两剖析样本是否来自同一总体。两样本的分界点通过求解分别代表两样本的级数模型曲线的交点而得。结果显示,混杂样本剖析的G-C级数模型法应用于药物代谢多态性的分型实用、合理,值得进一步推广。
The Gram-Charlier series model method for the analysis of mixed samples is introduced and applied to realize the rational classification of drug metabolism polymorphisms. A mixed G-C series model with two samples is presented. By using the computer software with self-defined nonlinear model fitting function, a nonlinear least square estimation of the model parameters is proposed step by step with the method of quasi-Newton. The theoretical frequency and actual The goodness-of-fit test of frequency between the two is reasonable. The rank sum test is used to infer whether the two analyzed samples come from the same population. The demarcation point between two samples is obtained by solving the intersection points of the curve models representing the two samples respectively. The results showed that G-C series model of mixed sample analysis applied to the classification of drug metabolism polymorphism is practical and reasonable, which is worth further promotion.