基于顺式调控元件的个体化基因表达预测模型及在免疫响应中的应用

Understanding the impact of cis-regulatory variants on gene expression is a fundamental step towards deciphering their functional roles in human disease. Substantial progress has been made on detecting statistical relationships between single-nucleotide variants and gene expression. But this approach has difficulties to identify functional variants, to consider the interaction within multiple variants, and to estimate the impact of external environment. In previous work, we developed novel personalized gene expression models based on cis-regulatory elements (CRE) to confer the the functional roles of cis-regulatory variants. In this proposal, we will continue to improve the methodology of our approach and to extend the utility of our models in genome analysis. We will build new models to investigate the contribution of interaction between variants and interaction betwwen genotype and environment. Then, models will be trained on multiple tissues and multiple types of immune response to enable wide usage for disease-related variant analysis. Finally, we will build an online database to present the impact prediction of all know variants in human genome. This project will advance our capability to interpret fuctional roles of cis-regulatory variants and to identify key CREs and genes in disease and immune response, representing a key step towards precision medicine.

解析基因突变对基因表达的影响是精准医学中的关键问题之一。已有研究基于SNP位点对基因表达量进行建模预测，但存在着难以确定有功能的突变、忽略突变之间的组合作用、未考虑外部环境影响等关键问题。近期我们提出了基于顺式调控元件（CRE）功能强度的个体化基因表达预测模型，为解析调控突变（CRE内突变）的功能提供了新方法。本项目将从方法及应用两个层面继续展开工作。方法层面上，进一步验证CRE内调控突变之间的组合作用、免疫响应时调控突变与外部环境的相互作用对个体基因表达的影响，并利用先验知识提高模型精度。应用层面上将拓展模型的应用范围，在多种组织、多种免疫响应条件下训练模型，将模型应用于疾病相关突变的分析中；最后建立在线数据库，提供人类基因组常见突变的功能预测结果。本项目的成功开展将为调控突变的功能解析提供新的计算方法，促进免疫响应中关键CRE和复杂疾病易感基因的挖掘，具有重要的科学与应用价值。

解析调控区基因突变对基因表达的影响是精准医学中的关键问题之一。已有研究基于临近基因的突变位点对基因表达量进行建模预测，但存在着难以解析突变的具体功能、忽略远距离突变与未考虑外部环境影响等关键问题。本项目在申请人已开发的基于顺式调控元件功能强度的个体化基因表达模型的基础上, 进一步在方法及应用两个层面开展了工作。方法层面上，利用先验知识(染色质表观遗传状态与三维互作状态)提高模型精度, 挖掘了免疫响应时调控元件与外部环境的相互作用对基因表达的影响。应用层面上将拓展模型的应用范围，在多种组织、多种免疫响应条件下训练模型，将模型应用于疾病相关突变的分析中；最后建立在线数据库，提供人类基因组常见突变的功能预测结果。本项目的成功开展将为调控突变的功能解析提供新的计算方法，促进免疫响应中关键调控元件和复杂疾病易感基因的挖掘，具有重要的科学与应用价值。