基于生物网络的人类疾病基因识别算法研究

项目名称： 基于生物网络的人类疾病基因识别算法研究

项目编号： No.61472133

项目类型： 面上项目

立项/批准年度： 2015

项目学科： 计算机科学学科

项目作者： 汤希玮

作者单位： 湖南第一师范学院

项目金额： 83万元

中文摘要： 计算方法致力于从大量候选疾病基因中识别最有希望的候选者，并将其提供给下游的筛选、验证实验或功能研究。尽管计算方法在过去几年有了很大改善，但大的挑战仍然是寻找与疾病真正相关的新基因以及证明新发现的基因与疾病之间的关联性。 本项目的研究基于一个被广泛认可的假定，即疾病的表型很少是单个受影响基因产物的结果，而是反映了复杂网络中相互作用的各种病理生物过程。本项目将整合蛋白质相互作用数据、转录因子绑定数据、基因表达数据、疾病表型数据、蛋白质结构域数据、基因本体注释信息等多源信息，研究不同类型的生物网络如疾病反应动态网络、加权相互作用网络、表型组-相互作用组网络等的构建技术。以这些网络为基础，进一步探究蛋白质之间的相互作用（即网络中的边）与疾病的关联关系、蛋白质复合物和疾病之间的关联关系。最后，从已知疾病基因出发，提出相应算法，识别新的与甲型流感H1N1、哮喘、乳腺癌和2型糖尿病密切相关的基因。

中文关键词： 疾病基因；多源信息融合；疾病反应网络；加权生物网络；蛋白质复合物

英文摘要： Computational approaches aim to identify the most promising genes or proteins among a larger pool of disease gene candidates. Their goal is to maximize the yield and biological relevance of further downstream screens, validation experiments or functional studies by focusing on the most promising candidates. Although computational methods have greatly improved in the past few years, the major challenge is still the discovery of novel genes that are involved in the disorder and the evaluation of the correlationship between the identified gene and disease. A widely held hypothesis underlying this project is that a disease phenotype is rarely a consequence of an abnormality in a single effector gene product, but reflects various pathobiological processes that interact in a complex network. This progect will intergrate multi-source information including protein interaction data, transcription factor (TF) binding data, gene expression data, disease phenotype data, protein domain data and gene ontology (GO) annotation,and research the constructing methods of different biological networks such as the dynamic networks involved in disease progression, the weighted interaction network, the phenome-interactome network. Based on these networks, the association between the interaction (edge) and disease will be modeled, and the correlationship between the protein complex and disease will be explored.In terms of the known disease gene, the corresponding algorithms are proposed to predict novel genes causing H1N1 influenza, asthma, breast cancer and type 2 diabetes.

英文关键词： disease gene;multi-source information fusion;disease response network;weighted biological network;protein complex