Administrative Supplement: Integrative analysis of lung cancer etiology and risk

RFA：PA-18-842(非-CA-20-006) 标题：行政副刊：肺癌病因和风险的综合分析 项目摘要/摘要 在过去的二十年里，许多全基因组关联研究(GWAS)揭示了最常见的 疾病具有多基因结构，其中具有小遗传效应的多个遗传变异累积在一起 影响疾病的发展。统计学、流行病学和临床遗传学的进步使 以多基因风险评分的形式展示多基因风险特征在定义高风险个体方面的威力 和低风险的疾病。然而，到目前为止，大多数基因研究都集中在基因上。 欧洲人群的同质性研究，考虑到在欧洲人口中样本的可获得性有限 不同的血统，以及由于不同种群之间等位基因比例的差异造成的混淆。这 这意味着没有充分了解GWA在种族差异方面的情况。 这项提案的目标是利用遗传多样性来开发和完善PR，通过解决 在基因研究中未得到很好代表的非裔美国人群体的祖先多样性；以及 说明影响非洲人获得和参与机会差距的个人和系统因素-- 美国人参与肺癌基因组检测的多基因风险，参与结果返回，并吸收 肺癌多基因风险评分信息在临床护理中的应用。我们假设特定于祖先的和 疾病亚型特定的多基因模型可以极大地改进风险预测，以识别高风险到低风险 在疾病预防和管理方面的疾病发展的个人。这项建议 利用现有的大规模、良好的基因分型和表型数据集，使用特定于祖先的疾病 基于子集、链接不平衡得分回归和机器学习关联分析实现 这是一个“癌症健康差异”的原则，并提高了非裔美国人人口的预测准确性。 此外，确定影响肺癌基因组差异的个体和系统水平因素 多基因风险检测将为制定更有针对性的干预措施提供信息，以改善获得机会， 参与和参与，这不仅可以增强模型预测，而且还具有有益的作用 阐明改善非裔美国人参与癌症基因组的策略对下游的影响 测试。

RFA: PA-18-842 (NOT-CA-20-006) Title: Administrative Supplement: Integrative analysis of lung cancer etiology and risk Project Summary/Abstract Over the past two decades, many genome-wide association studies (GWAS) have revealed that most common diseases have a polygenic architecture, wherein multiple genetic variants with small genetic effect cumulatively impact disease development. Advances in statistical, epidemiological, and clinical genetics enable to demonstrate the power of polygenic risk profiles in form of polygenic risk scores to define individuals at high- and low-risk of disease. However, most genetic research carried out to date has focused on genetically homogeneous studies from European populations given the limited availability of samples in populations of diverse ancestry and due to confounding from variability in allelic proportions among diverse populations. This implies that GWAS in ethnic disparities are not fully understood. The goals of this proposal are to leverage genetic diversity to develop and refine PRS by addressing the ancestral diversity in African American population that are not well-represented in genetic research; and to elucidate individual and system-level factors affecting disparities in access and participation of African- Americans in lung cancer genomic testing for polygenic risk, engagement with return of results, and uptake of lung cancer polygenic risk score information in clinical care. We postulate that the ancestry-specific and disease subtypes-specific polygenic models can greatly improve risk prediction to identify high- to low-risk individuals of disease development in terms of prevention and management of the disease. The proposal capitalizes on large-scale existing well-genotyped and phenotype dataset using ancestry-specific, disease subset-based, linkage disequilibrium score regression, and machine-learning association analysis to achieve the a “cancer health disparities” principle and to increase prediction accuracy in African American population. In addition, identifying the individual and system –level factors affecting disparities in lung cancer genomic testing for polygenic risk will inform development of more targeted interventions to improve access, participation, and engagement, which could not only enhance model prediction but also have salutary downstream impact by elucidating strategies to improve African-American engagement in cancer genomic testing.