大肠杆菌内源性转录终止子的支持向量机预测方法  

Prediction of Intrinsic Transcription Terminators in Escherichia coli by Using Support Vector Machine

作　　者：杜耀华[1] 王正志[2] 吴太虎[1] 

机构地区：[1]军事医学科学院卫生装备研究所,中国天津300161 [2]国防科技大学机电工程与自动化学院,中国湖南长沙410073

出　　处：《生命科学研究》2012年第6期471-478,495,共9页Life Science Research

基　　金：国家自然科学基金资助项目(60471003)

摘　　要：内源性转录终止子的计算预测是基因转录调控研究的重要内容,但当前方法的预测特异性偏低.在深入分析大肠杆菌内源性终止子中RNA发夹结构和多聚胸腺嘧啶区域等特征信号的基础上,为内源性终止子建立了一个由5个特征变量组成的包含序列组分、局部构象和能量分布信息的特征集,并根据此特征集实现了一种基于支持向量机的内源性终止子计算预测方法.针对大肠杆菌内源性终止子数据集和编码区阴性对照集的六重交叉验证测试证实了预测方法的有效性,对已知数据的预测平均正确率达到了99.4%.在对大肠杆菌全基因组限定范围内的搜索中,该预测方法可以成功地识别出绝大多数已知内源性终止子,与其他几种常用方法相比,预测结果总数大幅度减少,预测的特异性有了明显提高.Computational prediction of intrinsic transcription terminators is an essential task in the research of transcription regulation of gene,but the performances of current methods are still unsatisfying.According to the thorough analysis about signals such as RNA hairpin structure and T-region in E.coli intrinsic termi-nators,a 5 variable-included feature set which combines the sequence content,local conformation and ener-gy distribution information is selected.Based on such feature set,a new prediction method using support vector machine for intrinsic terminators is proposed.The favorable performance is achieved in 6-fold cross validation test on E.coli positive and negative control datasets,and the average prediction accuracy is 99.4%.This method is then used to scan the putative intrinsic terminators in the whole genome of E.coli.Comparing with several other methods,the total number of scanning hits decreases greatly when most of known intrinsic terminators are retrieved.The specificity of prediction results has been improved effectively.

关 键 词：转录终止 内源性终止子 RNA发夹结构 多聚胸腺嘧啶区域 支持向量机 

分 类 号：Q615[生物学—生物物理学]