中文摘要：

在生存分析中，Cox比例风险模型和加速失效模型是研究控制数量性状位点（QTL）的生存时间的常用的方法。本文首先借助于Cox比例风险模型给出了半参数的区间定位方法去表征生存性状的印记数量性状位点（iQTL）。此方法的明显优势是无需估计复杂的“讨厌”的基线风险函数。然后构造偏似然函数并根据其推断出iQTL的参数估计。采用期望最大化算法求解iQTL位置和遗传效应的极大似然估计值。所给统计推断方法保证了整个基因组范围内的显著iQTL的估计和检验，在一系列的零假设下，给出了所检测到的iQTL的印记模式。最后，此方法应用于分析一个公开发表的小鼠模型系统数据集。

英文摘要：

In survival analysis, Cox proportional hazard model or accelerated failure time model was the natural choice for linkage and association analyses of survival time with QTL/marker. A semi-parametric interval-based approach to （i）QTL detecting for survival traits with the aid of Cox PH model without estimating the complicated baseline hazard function which was ＂nuisance＂ was developed. A partial likelihood was obtained,leading to inference about the （i）QTL parameters and EM algorithm was given for solving the maximum likelihood estimates of iQTL location and effects. With this efficient procedure,a genome-wide estimating and testing the effects of significant （i）QTL were ensured and the imprinting patterns of the detected iQTL were statistically inferred under a series of null hypotheses. An application to a published dataset from a mouse model system was provided to illustrate the proposed approach.