癌症hallmark挖掘对于癌症的研究具有重要的意义，且癌症特定表型hallmark（本课题指参与到多种癌症转移中的功能基因集合）中的基因更可能是癌症转移中的驱动因子，但癌症hallmark的挖掘方法还很缺乏。本项目提出了一种从多种癌症的DNA甲基化交互网络中挖掘癌症hallmark的方法，并将hallmark应用于癌症转移biomarker的筛选。具体研究：(1)集成多种癌症大规模样本的DNA甲基化数据来为每种癌症构建DNA甲基化交互网络；(2)基于多种癌症的DNA甲基化交互网络来挖掘其中保守、且与转移有关系的社区作为癌症的hallmark；(3)基于癌症hallmark，筛选癌症转移biomarker并构建预后模型。我们的方法可以用于挖掘多个生物过程中的共同驱动因子，课题挑选出的biomarker可作为药物研发的潜在靶标，预后模型可为癌症病人的治疗提供参考。

Identifying the hallmarks of cancer is essential for cancer research, and the genes involved in the cancer hallmarks specific to a phenotype (hallmarks in this project means the functional gene sets involoved in metastasis processes of several kinds of cancer) are more likely to be cancer drivers of this phenotype. However, there is still no suitable method to solve this problem. Here, we propose a frame to identify the hallmarks based on the DNA methylation interaction networks of several kinds of cancers, and apply the hallmarks to select biomarkers of cancer metastasis. The main contents of this project are as follows: (1) For each cancer, a DNA methylation interaction network is constructed by integrating the large amount of samples’ DNA methylation data; (2) Based on the DNA methylation interaction networks of all the cancers, we propose a method to mine the conservative communities which are related to cancer metastasis across these networks and the conservative communities are set as cancer hallmarks; (3) Using the hallmarks, we select the biomarkers for cacner metastasis and construct prognostic model to predict cancer prognosis for cancer patients. In this project, we would propose a new frame to identify the common driver factors of several biological processes based on the according biological networks. In addition, the biomarkers may be target candidates for cancer drugs and the prognostic model can provide reference for individual therapy.