ESTIMATION & TEST OF MARKER ASSOCIATION IN FAMILY DATA

目前有越来越多的多态基因 可用于研究标记与人类关联的标记 疾病特征。对这类关联的了解提供了一个基本的 去做遗传咨询。因为对家族性疾病的研究 分离和连锁分析需要收集数据 在家庭和血统成员上，这种情况越来越多 从已有数据推断标记关联所必需的 在考虑到这些多重目的的情况下收集。因此， 全面、严谨、实用的统计方法 将开发用于从扩展的人类谱系数据进行分析 遗传标记与性状性状的关系 家族性疾病。不同的定性方法 性状、质量性状和发病年龄不同的性状 将会被开发出来。在数量性状的情况下，通常 常态假设将会放松。这些型号将允许 家系个体之间灵活的相关结构。 统计性质，如偏倚、一致性、功率和 估计量的稳健性与似然比检验 开发将通过数学分析和计算机进行研究 模拟。扩展到不同类型的确定将 被开发出来。这些方法将被应用于家系数据的分析 从各种正在进行的研究中获得，以确保 方法与真实数据相关。

There is currently an ever-growing number of polymorphic genetic markers available for the study of marker association with human disease traits. Knowledge of such associations provides a basic for genetic counseling. Because the study of familial disease by segregation and linkage analyses requires the collection of data on family and pedigree members, it is becoming more and more necessary to infer marker associations from data that have been collected with these multiple purposes in mind. Therefore, comprehensive, rigorous and practical statistical methodologies will be developed for analyzing from extended human pedigree data the associations between genetic markers and traits characteristic of familial diseases. Separate methodologies for qualitative traits, qualitative traits, and traits with variable age of onset will be developed. In the case of quantitative traits, the usual assumption of normality will be relaxed. The models will allow for a flexible correlation structure among the pedigree individuals. The statistical properties such as the bias, consistency, power and robustness of the estimators and the likelihood ratio tests developed will be studied by mathematical analysis and computer simulations. Extensions to different types of ascertainments will be developed. The methods will be applied to analyse pedigree data obtained from various ongoing studies in order to be sure that the methods are relevant to real data.