Genetic Epidemiology of Breast Cancer: Risk, Instability, and Statistical Methods

描述(申请人提供)：乳腺癌是发病率和死亡率的主要来源，仍然是美国女性中最常见的癌症，2008年有超过18.3万例新病例(Ries等人)。2008年)。这项应用解决了乳腺癌发展的两个方面：种系突变，如两个肿瘤抑制基因BRCA1/2，以及以拷贝数变化和等位基因丢失为特征的体细胞突变。这项应用的第一个目的是从没有外部疾病发病率数据的基于人群的病例对照研究中估计像BRCA1/2这样的候选基因的年龄相关外显函数。从数据和建模框架获得的估计将被用来估计突变携带者的概率，并预测女性的乳腺癌风险。一个共同的脆弱性模型将被用来解释广泛观察到的家庭之间的实质性风险异质性。该模型将扩展到一般的多变量脆弱模型，以允许不同类型的关系和多个候选基因相关表型的不同程度的相关性。尽管最近在确定乳腺癌新的候选基因方面取得了成功，但大约四分之三的乳腺癌病例在家庭中没有聚集性。人们认为，癌症的发展是由于染色体位置上的遗传异常积累的结果，这些染色体位置对维持正常的细胞功能至关重要。来自特定染色体位置的遗传信息的丢失和获得被认为是肿瘤抑制基因或癌基因参与肿瘤发生的迹象。利用基因组不稳定性数据，这项应用的第二个目标是识别与乳腺癌肿瘤发展有关的新的致癌网络。基于图形模型的方法将通过使用新的稀疏回归和多重测试技术来开发。随着高通量基因分型技术的广泛应用，研究人员能够进行大规模的全基因组关联和基因组不稳定性研究。这使我们能够识别可能对乳腺癌风险只有温和影响的新的候选基因座，或者在这样一个以前不可行的细节上发现致癌网络。迫切需要对这些新基因进行表征，并为这些网络提供推断。这里提出的方法就是朝着这个方向努力的。公共卫生相关性：该项目旨在表征与人群中基因相关的乳腺癌风险，估计突变携带者的概率，并根据这些基因的突变状态预测健康个体患乳腺癌的风险。它还旨在获得关于基因组不稳定性在乳腺癌发生和发展中的作用的知识。最终目标是为设计有效的针对个人的预防策略提供洞察力，以降低乳腺癌风险。

DESCRIPTION (provided by applicant): Breast cancer is a major source of morbidity and mortality, and remains the most common cancer occurring in U.S. women, with over 183,000 new cases for 2008 (Ries et al. 2008). This application addresses two aspects of breast cancer development: germline mutations, as in two tumor suppressor genes BRCA1/2, and somatic mutations characterized by copy number changes and allelic loss. The first aim of this application is on the estimation of the age-dependent penetrance function of candidate genes like BRCA1/2 from population-based case-control studies without external disease incidence data. The estimates obtained from the data and the modeling framework will be used to estimate the mutational carrier probability and predict the breast cancer risk of a woman. A shared frailty model will be used to account for the widely observed substantial risk heterogeneity among families. The model will be extended to a general multivariate frailty model to allow for varying degrees of the correlation for different types of relations and multiple candidate gene-related phenotypes. Despite recent successes in identifying novel candidate loci for breast cancer, about three- quarter of breast cancer cases having no clustering in families. It is believed that cancer develops as a result of an accumulation of genetic aberrations at chromosomal locations that are critical in maintaining normal cell functions. The loss and gain of genetic information from specific chromosomal locations are considered an indication for the involvement of tumor suppressor or oncogenes in the tumorigenesis. Using genomic instability data, the second aim of this application is on identifying novel oncogenic networks involved in breast cancer tumor development. Graphical model-based methods will be developed by using novel sparse regression and multiple testing techniques. As the high throughput genotyping technologies become widely available, researchers are able to conduct large scale genome-wide association and genomic instability studies. This allows us to identify novel candidate loci that may have only moderate effect on breast cancer risk or discover oncogenic networks at such a detail that is not previously feasible. There is a great need to characterize these novel genes and provide inference for these networks. The methods proposed here are efforts towards this direction. PUBLIC HEALTH RELEVANCE: This project aims to characterize the breast cancer risk in relation to genes in the population, estimate the mutation carrier probabilities, and predict a healthy individual's risk for developing breast cancer given the mutational status in these genes. It also aims to gain knowledge on the role of genomic instabilities in breast tumor initiation and progression. The ultimate goal is to provide insight in devising effective individual-tailored prevention strategies for reducing breast cancer risk.