人类复杂疾病相关的非编码区调控性单核苷酸变异预测研究

Human complex diseases are determined by multiple genetic loci, environment factors and other elements additively. The widely applied approaches to analyze genetic associations of complex diseases include genome wide association study (GWAS) and resequencing of individual genomes, while the former approach receives more acceptance with the advantages of high efficiency and low cost. GWAS experiments rely on other technologies or methods to detect the exact causal loci in diseases, such as the costly targeted sequencing technology or bioinformatics methods. For single nucleotide variations, most bioinformatics approaches aimed to annotate the coding region loci which may affect protein structures and functions; while a few approaches considering the non-coding region loci did not count either the regulatory mechanisms or the quantization of regulatory capability. Given the fact that most GWAS loci located in the non-coding region which accounted for 98% of human genome, this project will develop bioinformatics approaches applied in elucidating non-coding genetic association of complex diseases, to quantitatively analyze the potential regulatory mechanisms of single nucleotide variation in transcriptional regulation, splicing regulation and post-transcriptional regulation, targeting prediction of single nucleotide variations associated to complex diseases. Achievements of this project will advance the large-scale genetic association analysis and ultimately accelerate the study of complex diseases.

人类复杂疾病受基因组范围内多遗传位点、环境及其他因素的综合影响。广泛应用的复杂疾病遗传关联性研究手段有全基因组关联性分析（GWAS）与基因组重测序分析，其中GWAS以高效率、低花费的优点应用更为普遍。GWAS技术依赖于其他研究手段与方法帮助进一步确定疾病的确切致病性遗传位点，如昂贵的靶向测序技术或者生物信息学方法。针对单核苷酸变异，已有的生物信息学方法大多集中于注释可能影响蛋白质结构与功能的编码区位点；少数分析非编码区位点功能的方法要么无法考虑其潜在的调控机制，要么不能量化其调控功能大小。由于GWAS位点大多分布在非编码区域，且非编码区域占人类基因组约98%，本项目拟针对复杂疾病在非编码区的遗传关联性问题设计生物信息学方法，分析单核苷酸变异位点可能的转录调控、剪切调控与后转录调控机制，定量预测复杂疾病相关的调控性单核苷酸变异。该研究将推动复杂疾病的大规模遗传关联性分析，加速复杂疾病的研究。

与疾病相关的调控性单核苷酸多态性研究旨在寻找除编码区遗传变异位点之外的其他靶向位点，对于推动临床健康研究具有重要意义。该方向传统研究方法存在靶点识别分辨率低、实验技术花费高、实验过程时间长等不足。本项目充分组织各类分子生物学数据与数据库资源，依据基因的调控机理设计了融合公共数据资源并快速预测疾病相关单核苷酸多态性的多个工具。同时，依据单核苷酸多态性带来的数据关联不准确性，将项目研究延伸至如何将其运用于辅助测序数据回帖以改善此不确定性。主要研究成果包括以下三个方面，并以学术论文的方式发表：开展了以单核苷酸多态性为主，同时兼具其他多类基因组序列变异的，变异分子功能、表型功能的预测与可视化平台研究；开展了在肝癌细胞中针对单核苷酸多态性对增强子的调控功能预测研究；开展了针对单核苷酸多态性对测序数据回帖影响的研究，从而从数据层面提高下游研究中单核苷酸多态性与疾病关联的预测准确性。这些研究的开展完成并延伸了项目预期目标，提出了新的生物数据资源的融合方法，优化了突变位点对疾病数据关联的影响，设计了新的疾病靶点快速预测与可视化平台，从生物信息学角度发展并丰富了疾病的分子机理，对于临床健康研究具有重要意义。