Statistical Issues in the Design and Analysis of Studies of Genetic Varians HT ..

项目D：遗传学研究设计和分析中的统计问题 变种，HT和癌症。病例对照研究正变得越来越复杂， 目前正在同时评估暴露和宿主因素与风险的关系。一个 现在，大量的努力集中在描述人类基因组的变异上， 可能与常见疾病的风险有关。这包括建立广泛的SNP 数据库和（最近）一个雄心勃勃的项目，旨在了解模式的联系 通过开发人类基因组单倍型图谱（ HapMap）。 本项目的主要重点是进一步发展与统计方法有关的 使用不相关受试者的病例对照研究来评估候选基因可能 在癌症风险和个体对激素反应的差异性中发挥作用 具体来说，暴露。此应用程序中的所有其他计划项目将评估角色 候选基因在生殖相关癌症中发挥作用;在许多方面，P01的项目 是“下一代”关联研究的原型， 从HapMap中提取。我们在项目的这一部分中考虑的问题是预期的 适用于所有此类研究。 该项目的第二个目标是重点发展和拟订统计数据， 处理暴露测量误差对风险估计的影响的方法， 尤其是在巢式病例对照研究中。我们的注意力将集中在其他计划的需求上。 项目，我们计划最初执行的测量误差的影响进行仔细的分析， 在项目D中提出的问题上，我们将继续发展我们的 对流行病学研究的测量误差校正也有更广泛的兴趣。

PROJECT D: Statistical Issues in the Design and Analysis of Studies of Genetic Variants, HT and Cancer. Case-control studies are becoming increasingly complicated as exposures and host factors are being concurrently evaluated for their relationship to risk. An extraordinary amount of effort is now focused on describing variation in the human ge.nome which may be relevant to the risk of common diseases. This includes the establishment of extensive SNP databases and (very recently) an ambitious project designed to understand the patterns of linkage disequilibrium of SNPs through the development of a haplotype map of the human genome (the HapMap). The primary emphasis in this Project is the further development of statistical methods related to the use of case-control studies of unrelated subjects to assess the role that candidate genes may be playing in the risk of cancer generally and in the individual variability of response to hormone exposures specifically. All of the other planned projects in this application will be evaluating the role that candidate genes play in hormonally related cancer; in many respects the projects of this P01 are proto-typical of the "next generation" of association studies that are envisioned to be resulting shortly from the HapMap. The issues that we consider in this portion of the project are anticipated to be applicable to the full range of such studies. The second aim of this project is focused upon development and elaboration of statistical methods for dealing with the effects of exposure measurement error upon risk estimation, particularly in nested case-control studies. Our attention will be on the needs of the other planned projects and we plan initially to perform a careful analysis of the effects of measurement error in mammographic density upon the problems raised in Project D. We will continue to develop our interests in measurement error correction for epidemiological studies more broadly as well.