基于网络拓扑相似性预测潜在致病基因

发布时间：2018-06-01 02:11

本文选题：网络结构 + 二部图　；参考：《安徽大学学报(自然科学版)》2017年05期

【摘要】：相关疾病基因的发现和预测是人类基因组研究的重要目标.近些年,一些研究者通过基于网络结构的方法来解决这个难题.然而,大多数方法在推理过程中仅使用了局部的网络信息,并且仅限于推理单一基因的关联.并且这些方法很少考虑到疾病-基因关联网络的网络拓扑性.笔者提出一种改进的基于二部图网络结构推理(improved network-based inference)的计算方法.该方法基于已知的疾病-基因网络拓扑相似性来发现更多潜在致病基因.文中使用的是OMIM数据库中的203种疾病的数据,通过留一交叉验证法验证实验,并获得了88.9%的AUC值.与文中提到的另外两种方法相比,该文方法能够有效地预测潜在致病基因.
[Abstract]:The discovery and prediction of related disease genes is an important goal in the study of human genome. In recent years, some researchers have solved this problem through a network based approach. However, most of the methods use only local network information in the reasoning process, and are limited to the association of single genes. Considering the network topology of the disease gene association network, the author proposes an improved method based on the two part graph network structure reasoning (improved network-based inference). This method is based on the known disease gene network topology similarity to discover more potential pathogenic factors. The 203 species in the OMIM database are used in this paper. The data of the disease were verified by a cross validation method, and 88.9% of the AUC value was obtained. Compared with the other two methods mentioned in the article, the proposed method can effectively predict the potential pathogenic genes.
【作者单位】：安徽大学计算智能与信号处理教育部重点实验室;
【基金】：国家自然科学基金资助项目(61172127)
【分类号】：R319;TP301.6

【相似文献】