基于DNA甲基化波动性挖掘泛癌基因组范围的动态甲基化特征及功能分析

DNA methylation is a dynamic and reversible epigenetic modification that plays an important role in cell differentiation and tumorigenesis by participating in transcriptional regulation. High-throughput omics data in multi- tissue and cancer types have been constantly accumulated and facilitates the genome-wide mining and functional analysis of the dynamic methylation characteristics. This project is carried out according to the following process: high-throughput data integration → dynamic methylation feature mining → functional analysis. Benefit from the technology of high-speed file index system, the pan-cancer wide whole genome high-throughput omics data have been integrated and made fast accessible. Next, based on the wavelet theory and statistical tests, a DNA methylation volatility model is constructed for the quantitative assessment of methylation volatility among normal/cancer tissues. Then, we establish a landscape for the dynamic changes of the methylation patterns in the cancer genomes. The pan-cancer wide and inter-tumor specific dynamic methylation characteristics then can be identified based on the measurement of the wavelet relative entropy. Further, we construct a dynamic methylation-mediated abnormal transcriptional regulatory network and try to reveal the molecular mechanism in which the dynamic methylation characteristics affecting the co-localization of transcription factors and abnormally assembling the tumor-specific super enhancers that regulate the key tumor factors by a strategy of combining regulatory network analysis and experimental verification. The dynamic methylation characteristic screened in this project is an important complement to the tumor heterogeneity system and should be of great significance for understanding the tumor heterogeneity and provide an important reference for the further study of tumor broad-spectrum therapeutic targets and cancer type specific prognosis markers.

DNA甲基化是一种动态可逆的表观遗传修饰，通过调控基因表达参与细胞分化和肿瘤发生。多组织、癌症类型中高通量组学数据的产出促进基因组范围动态甲基化特征的挖掘及功能分析。本项目依照高通量数据整合→动态甲基化特征挖掘→功能分析的研究思路，采用高速文件索引系统整合泛癌基因组高通量多维组学数据；基于小波理论和统计学检验构建DNA甲基化波动性模型，在正常及癌症组织间定量评估甲基化波动性，绘制癌症基因组甲基化模式的动态变化图谱，利用小波相对熵测度筛选泛癌水平以及癌症间特异性动态甲基化特征；构建动态甲基化特征介导的异常转录调控网络，通过网络分析和实验验证相结合的策略，揭示动态甲基化特征影响转录因子共定位、异常装配肿瘤特异超级增强子、调控关键肿瘤因子的分子机制。本项目筛选的动态甲基化特征是对肿瘤异质性相关分子特征体系的重要补充，对深入理解肿瘤的异质性和筛选肿瘤广谱治疗靶点及特异性靶向预后标记具有重要意义。

恶性肿瘤是对人类危害程度极高、异质性极强的复杂疾病之一，与众多遗传 和表观遗传因子的功能异常显著相关。DNA甲基化在肿瘤的形成和发展、基因 组印记的过程中起到关键的作用，抑癌基因的高甲基化和原癌基因的低甲基化是 癌细胞中显著的甲基化变异模式。本项目基于DNA甲基化水平动态变化的特点，利用波动性理论对CpG位点的甲基化水平在组织间波动的程度进行了定量评估，分类基因组CpG位点为三类：组成性DNA甲基化模式（ConCG）、波动性DNA甲基化模式（WavCG）和差异DNA甲基化模式（DiffCG）位点。挖掘、筛选了泛癌症水平和独立癌症类型中DNA甲基化模式发生急剧变异的CpG位点（称为甲基化"Quake"位点）。其中20个CpG位点在泛癌症水平上发生了“ConCG→DiffCG”的甲基化模式突变，能够对癌症类型进行区分，且与转录因子的活性显著相关。另有51个CpG位点在泛癌症水平上发生了“DiffCG→ConCG”的甲基化模式突变，呈现出组织类型特异的聚类模式。通过小波相对熵算法构建了在独立癌症类型中筛选甲基化“Quake”位点的模型，用于探讨泛癌症水平的癌症间一致性与异质性。在16种癌症中筛选得到42个甲基化模式突变的CpG位点在3种或3种以上的癌症类型中具有一致性，其中cg16648632在10个癌症类型中共甲基化模式突变；184个CpG位点为15个实体肿瘤异质性甲基化模式突变位点，其中有11个异质性甲基化模式突变基因为已知的印记基因。 预后分析显示，癌症特异的甲基化模式突变CpG位点中19个位点与肿瘤不良预后显著相关，且超过47%的预后相关的肿瘤特异甲基化模式突变位点落于超级增强子的作用范围内。

内容获取失败，请点击重试