circRNA具有多种生物功能，如作为microRNA海绵、调控其它ceRNA或与RNA结合蛋白(RBP)相互作用，某些circRNA还具有编码蛋白的能力。这些功能都潜在影响着复杂疾病的发生发展过程，然而目前对于circRNA具体的作用机制尚不清楚，与疾病的关联关系也很难被鉴定。本项目拟结合RNA甲基化修饰特征，构建基于深度学习的circRNA编码潜力预测模型，并通过构建综合的ceRNA调控网络，阐释circRNA对于其它ceRNA的调控关系，同时利用蛋白质相互作用(PPI)等信息，研究潜在的circRNA与RBP作用关系及其动态特性。在获得cricRNA的各种潜在特性、调控作用关系后，利用机器学习、数据挖掘及智能计算等方法预测circRNA与复杂疾病之间的关联关系，解释其作用机理，以期为复杂疾病的精准诊断提供理论基础和方法依据。

CircRNAs possess a variety of biological functions. For examples, they can act as microRNA sponges, regulate other ceRNAs or interact with RNA-binding proteins (RBP). Some circRNAs may function in the same ways as proteins. All these functions potentially affect the occurrence and development of complex diseases. However, many specific functional mechanisms are still unclear, and it is still very challenging to identify the associations between cricRNAs and complex diseases. This project intends to combine the characteristics of RNA methylation modifications to construct a deep learning based model for predicting circRNA coding potentials. By constructing integrated regulatory networks between ceRNAs involved in circRNAs, to elucidate the regulatory relationship of circRNAs to other ceRNAs. At the same time, with the information of protein-protein interactions(PPIs), we will study the potential relationships between circRNAs and RBPs and their dynamic characteristics. After obtaining the various potential characteristics, regulatory and functional relationships of cricRNAs, we will design an array of machine learning, data mining and intelligent computing methods to predict and understand correlations between circRNAs and complex diseases. The project is expected to provide a theoretical and methodological basis for the precise diagnosis of complex diseases.