基于复杂网络理论的2型糖尿病中医证治规律及其机制研究

本项目在复杂性科学相关理论的指导下，以开放的组学数据库和自主研发的中医药诊疗支撑平台为依托，以复杂网络建模研究为主线，从系统、器官、组织细胞及分子等多个层面，梳理中医理、法、证、方、药的密切关系和演变规律，探索2型糖尿病中医证治规律及其作用机制。选择多家中医医院2型糖尿病治疗的病历资料（包括电子病历），抽取特征信息作为复杂网络构建的临床数据源，分析网络的分布规律和中心性特征，运用随机森林、logistic回归、贝叶斯网络等方法构建指标预警模型。在医生用药、证候演变、指标变化和临床表型变化等方面，通过比对2型糖尿病复杂网络模型模块识别和动态变化规律，在研发特定算法的基础上，构建基于复杂网络关联规则的糖尿病数据库；最后，运用计算机模拟筛选方法，结合开放的组学数据库等资源，探索医生处方用药的证据关联，找出临床有效方药的治疗靶点，为阐明临床有效方药的药效物质基础和降糖机制提供新方法、新路径。

Under the guidance of the complexity science theory and the project elaborates the close relationship of medical knowledge ,therapy, and prescription based on the open genomics database and TCM clinic support platform. Simultaneously, it also probes type 2 diabetes treatment of TCM and mechanisms of drug treatment. We select the electronic medical records of several Chinese medicine hospital , extract feature information as clinical data sources to build complex networks, and use select random forest, logistic regression, Bayesian networks and other methods as as the basis for constructing the early warning models. In terms of the clinical medicine, syndromes evolution, index changes and clinical phenotype, it plan to build a database, which is based on the complex networks of association rules by contrasting dynamic changes of the network module. Lastly, We use the computer simulation method for screening, combine open genomics database resources, explore the related evidence of the prescription drugs, and find the therapeutic targets of the clinically effective prescription. These tasks can provide a new method for clarifying the clinical efficacy of the material basis of effective herbs and the potential hypoglycemic mechanism.