Analysis of Genetic-Environment Networks in Spondyloarthritis

大多数表型变异，包括那些涉及复杂疾病，如强直性脊柱炎 (AS)和药物反应的差异，是由多种遗传和 环境因素现有的方法可以提供分析工具，但随着 HapMap项目提供了一个涵盖数百万种SNP和单倍型的综合目录， 人口和高通量基因分型技术的快速发展，遗传学的范式转变， 从个体标记分析到全基因组关联研究和系统水平的复杂性状研究 分析是必不可少的。复杂系统的全基因组关联研究和系统水平分析 疾病在三个方面提出了巨大挑战。首先，确保全基因组的 0.05的显著性水平用于使用传统统计方法测试数百万个SNP。二是多数 表型变异是多种遗传和环境因素综合作用的结果 通过基因之间以及基因和环境之间复杂的相互作用。检测交互 在基因或SNP标记之间进行比对是一项艰巨的任务。第三，大多数现有的分析方法分析每个标记 (or单倍型）和表型，并不考虑网络结构之间的多个 表型和多个标记。因此，需要提出新的技术来解决这些问题。 具有挑战性的任务。本项目的总体目标是（1）发展全基因组的非线性统计 关联研究以确保全基因组显著性水平，（2）开发新的统计方法， 基因相互作用的检测和构建基因相互作用的有效计算算法 网络，（3）开发多表型网络建模的概念框架，以及（4） 开发或采用新的统计方法，用于多表型和多标记的联合分析。作为 是一种复杂的疾病。来自项目1-3的基因型和表型数据，用于剖析复杂的遗传 AS的体系结构将用于方法学的开发和评估以及真实的数据分析。

Most phenotypic variations, including those involved in complex diseases such as ankylosing spondylitis (AS) and differences in drug response, are generated by integrated actions of multiple genetic and environmental factors. Existing methods may provide tools for analysis, but with the imminent completion of the HapMap Project providing a comprehensive catalogue of millions of SNPs and haplotypes across diverse populations and rapid development of high throughput genotyping technologies, a paradigm shift for genetic studies of complex traits from individual marker analysis to genome-wide association studies and systemslevel analysis is indispensable. Genome-wide association studies and systems-level analysis for complex diseases raise great challenges in three aspects. First, it is practically impossible to ensure a genome-wide significance level of 0.05 for testing millions of SNPs using traditional statistic methods. Second, most phenotypic variations are generated by integrated actions of multiple genetic and environmental factors through complex interactions between genes, and between gene and environments. Detecting interactions among genes or SNP markers is a daunting task. Third, most existing analytic methods analyze each marker (or haplotype) and phenotype individually, and do not consider network structures among multiple phenotypes and multiple markers. Therefore, new techniques need to be proposed to address these challenging tasks. The overall goal of this project is (1) to develop nonlinear statistics for genome-wide association studies for ensuring genome-wide significance levels, (2) to develop novel statistical methods for detection of gene interaction and efficient computational algorithms for construction of genetic interaction networks, (3) to develop a conceptual framework for network modeling of multiple phenotypes, and (4) to develop or adopt novel statistical methods for joint analysis of multiple phenotypes and multiple markers. AS is a complex disease. Genotype and phenotype data from projects 1-3 for dissecting complex genetic architecture of AS will be used for development and evaluation of methodology, and real data analysis.