利用相似显著性序列特征预测白血病疾病基因  

Predicting disease genes of leukemia by using similar significant sequence features

作　　者：袁芳[1] 周艳红[1] 何光源[2] 

机构地区：[1]华中科技大学湖北省生物信息与分子成像重点实验室,武汉430074 [2]华中科技大学中英联合实验室,武汉430074

出　　处：《华中师范大学学报（自然科学版）》2012年第4期461-467,共7页Journal of Central China Normal University：Natural Sciences

基　　金：国家自然科学基金重大研究计划(90608020;30971642)

摘　　要：研究发现新的白血病致病基因对揭示白血病机理,寻求有效治疗方法有重要意义.选取了mRNA和protein序列的330个物化特征,发现某些特征在五类白血病(T-ALL、ALL、AML、APL和JML)的已知疾病基因中表现出相似的显著性差异.基于这个特点设计了一种新的白血病疾病基因预测方法.测试结果表明该方法能够有效地从众多候选基因中预测出真实的白血病基因,效果明显优于其他预测方法.Finding new leukemia genes undoubtedly has important significance for revealing its pathogenesis and seeking its effective treatment. In this research we chose 330 physical and chemic features of mRNA and protein and found certain of these features showed similar significant differences on known disease genes of five kinds of leukemia （T-ALL, ALL, AML, APL and JML）. This property was used to develop a novel approach to predict leukemia disease genes. Test results demonstrate that this approach can effectively detect actual leukemia disease genes from numbers of candidate genes and the effect of this approach obviously exceeds other approaches.

关 键 词：白血病 序列特征 疾病基因预测 

分 类 号：R318.04[医药卫生—生物医学工程]