动态差异DNA甲基化驱动的泛癌风险因子识别及其作用机制研究

Disordered DNA methylation, as one vital abnormity of epigenetic modifications, has played a similarly important role to that of genetic instability. However, few studies have focused on the studies of dysfunctional molecular regulatory mechanisms and risk factors driven by abnormal epigenetic modifications, based on large-scale pan-cancer data input, which significantly delay the comprehensive decoding of human complex diseases. This project roots in the relatively sophisticated cellular and molecular regulation network, from the perspective of pan-cancer, based on multiple omics data integration of genome, epigenome and regulome, and information mining strategies, to systematically characterize cancer- and normal- specific molecular regulation networks, construct and analyze pan-cancer-related dysregulated molecular networks, further explore abnormal epigenetics-driven pan-cancer-related risk factors and their mechanisms of action, and dissect common and specific underlying molecular dysregulated mechanisms of pan-cancer. Finally, great efforts should also be contributed to the establishments of bioinformatic platform for data storage and visualization, to lay the foundation for comprehensive understanding of pan-cancer-related dysregulated mechanisms, and to provide novel approaches for prognosis and prioritized treatments of human cancers from the perspective of abnormal epigenetic modifications.

失调DNA甲基化，作为一种极其重要的表观修饰异常，俨然已被推到同基因组不稳定同等重要的位置上。然而，基于大规模泛癌数据的异常表观修饰驱动的分子失调机制研究，以及风险因子的识别却少之又少，大大阻碍了全面破译人类复杂疾病的进程。本课题以相对完善的细胞分子调控网络为基础，从大规模泛癌研究的视角，基于基因组，表观组和调控组的多维组学数据融合与信息挖掘策略，系统全面地刻画癌症及正常状态特异的分子调控网络，构建并解析泛癌相关的失调分子调控网络，进而深度挖掘异常表观驱动的泛癌相关风险因子及其作用机制，深刻剖析泛癌相关的普遍及特异的分子失调作用机制。最后搭建数据存储及可视化为一体的生物信息学服务平台，为实现系统全面地理解泛癌分子失调机制提供坚实的理论基础，更为实现从表观修饰异常角度的人类癌症的诊断开辟全新的途径和优化的治疗方案。

癌症是一种恶性的复杂疾病。其发生发展具有多阶段性、多时间点性；而其致病机理具有多样性，包括基因组和表观组层面（动态失调的甲基化修饰）的异常。为了更深刻地理解对于人类癌症的认知，整合多维度的组学分子调控数据，对不同种类的癌症进行尽可能全方位的刻画，成为实现人类攻克癌症计划的重要环节。本项目旨在完善并搭建潜在的分子调控网络，基于多组学整合，及多类型分子调控关系的整合，并利用分子的基因表达层面信息对其进行过滤，进而识别构建疾病状态特异的分子失调网络。这不仅帮助剔除特定疾病条件下未发挥功能作用的分子调控关系，同时也降低了生物信息学计算的高维度难度。后续由全基因组层面转变为集中于癌症关键自噬通路的研究，进一步对数据进行了降维处理。未来将关注于细胞凋亡、细胞衰老相关通路的研究比较分析。通过疾病之间的横向比较，识别疾病状态特异的分子驱动网络。整合临床预后生存信息，对识别的分子标记物，以及分子标记模块/通路进行系统性的筛查，有助于识别具有临床转化应用价值的高特异性的分子标记物，这对于不同癌症的早期诊断，以及病人的预后预测都是极有价值的。同时，这些潜在的分子标记物，可以作为新药的研发的突破口，将极大提高药物靶向治疗的精确度。