2型糖尿病发生发展过程的临界状态预测理论和生物信息学方法

本项目通过动力系统理论和网络生物学来研究二型糖尿病及其并发症的临界转化过程，建立基于小样本、高维、大噪声数据的临界状态检测理论和生物信息学方法，实现糖尿病临界状态的定量评估和预测。特别是，(1)开发针对糖尿病异源多组学数据的挖掘、融合、网络构建方法；(2)基于糖尿病高维多组学数据，发展小样本的网络熵方法，建立动态网络熵标志物；(3)发展糖尿病及其并发症临界状态识别方法和模型，分析糖尿病稳态失衡的关键节点和关键分子网络；(4)理论与实验相结合，通过研究动物模型等的糖尿病发生发展不同阶段多层次组学数据，揭示糖尿病临界状态关键基因群和其分子网络。为真正实现糖尿病及其并发症临界状态评估与预测（如，预防性和个体化医疗）提供生物信息学理论基础和方法支撑。

It is a nonlinear and dynamical process from heath state to diabetes involving critical transitions or tipping points at a network level. This project aims to develop a computational model to analyze such critical transitions based on both theoretical analysis and omics data, and further establish a bionformatics method for identifying the tipping points or predicting the diabetes. Specifically, (a) derive molecular networks before or after the critical transitions by integrating multi-level omics data; (b) identify dynamic network biomarkers to characterize the critical transition based on network entropy measure; (3) develop a prediction model and bionformatics method to identify the early signals of the critical transitions of the diabetes on the basis of single samples.