【摘要】 目的比较病例-同胞对照设计三种统计方法的检验效能。方法应用Monte-Carlo模拟方法,使用Slink软件产生的模拟家系资料,比较同胞-传递不平衡检验(s-TDT)、同胞组不平衡检验(SDT)和同胞对关联检验法(SAT)三种统计方法的检验效能。结果(1)SDT方法的I型误差最接近检验水准α=0.05;(2)在三种遗传模式中,s-TDT方法的效能最高,SAT方法的效能最低。结论病例-同胞对照设计使用患者的同胞作为对照,避免了人群分层现象,可以用来评价遗传因素与疾病之间的关联。更多还原

【Abstract】 Objective To compare statistical power of three statistical methods of case-sibling design.Methods Family data were simulated by Slink.Statistical power of three statistical methods of case-sibling design was compared by Monte-Carlo simulations.Results(1)the type I error of SDT was the most close to the nominal significance level(α=0.05);(2)Across a variety of genetic models,s-TDT and the SDT are generally more powerful than the SAT,and the SDT is generally less powerful than the s-TDT.Conclusion Case-sibling design avoids population stratification by using siblings of cases.So it might be used to evaluate association between genetic effects and disease.更多还原