融合基因关系网和功能预测恶性胶质瘤基因方法研究  

Research of Prediction of GBM Genes by Integrating Gene Relational Network and Gene Function

作　　者：孙红梅[1] 常志强[1] 张淑娟[2] 许艳[1] 

机构地区：[1]哈尔滨医科大学生物信息科学与技术学院,黑龙江哈尔滨150086 [2]哈尔滨医科大学基础医学院,黑龙江哈尔滨150086

出　　处：《现代生物医学进展》2014年第13期2549-2553,共5页Progress in Modern Biomedicine

基　　金：黑龙江省自然科学基金资助项目(D201116)

摘　　要：目的:建立挖掘恶性胶质瘤候选基因的方法并进行系统分析。方法:结合恶性胶质瘤已知通路内基因和发生点突变和拷贝数改变的基因构建扩展基因关系网络,计算并分别寻找在网络中度和中心性得分高,脆弱性为正数的节点(基因),将满足一种或多种测度并与已知恶性胶质瘤基因共功能的基因作为恶性胶质瘤候选基因。最后,通过文献验证方法评价多种测度预测恶性胶质瘤基因的效能。结果:融合基因功能后,利用基因在网络中的度和脆弱性可识别大部分恶性胶质瘤基因,但利用中心性预测的结果较差;当将三个测度融合后,效能并没比单独使用脆弱性高。结论:融合基因功能关系和网络脆弱性是预测恶性胶质瘤基因的最佳测度。Objective： To establish and analysis the method of mining malignant glioma candidate genes. Methods： Combining the pathway of malignant glioma, point mutations and copy number changes in genes, based on the gene function and degree, vulnerability and the centrality of genes in network, this study analysis and evaluate the effectiveness of these characteristics in the identification of malignant glial the effectiveness of tumor related genes. Results： Combining the gene function, the degree and vulnerability of the genes in the network can identify most of the malignant glioma gene, while central forecast is poor. When combining three measures, the performance did not perform better than the use of vulnerability. Conclusion： The gene functional relationships and network vulnerability is the best measure to predict malignant glioma genes.

关 键 词：基因关系网络 脆弱性 中心性 恶性胶质瘤 

分 类 号：R739.41[医药卫生—肿瘤]