论文部分内容阅读
目的发展快速的似然比罕见遗传位点关联性分析方法。方法在精确分布的基础上采用混合分布作为替代。通过模拟研究来比较混合分布和精确分布;以及通过GAW17数据来验证近似的和精确的似然比检验。结果不同的混合分布参数估计方法基本获得一致的估计值。GAW17数据分析中似然比检验根据混合分布和精确分布获得p值基本一致。结论混合分布不但有效提高了似然比罕见遗传位点关联性分析方法的计算速度,也同时保持了其高的统计效能。
Objective To develop rapid likelihood ratio analysis of rare genetic association. The method uses a mixture distribution as an alternative based on the exact distribution. The mixed distributions and exact distributions were compared by simulation studies; and the approximate and accurate likelihood ratio tests were verified by GAW17 data. The results of different hybrid distribution parameter estimation method basically get a consistent estimate. In the GAW17 data analysis, the likelihood ratio test is basically consistent with the p value obtained from the mixture distribution and the exact distribution. Conclusion The mixed distribution not only effectively improves the computational speed of the RLSR analysis but also maintains its high statistical power.