计算机与现代化 ›› 2012, Vol. 1 ›› Issue (11): 30-32.doi: 10.3969/j.issn.1006-2475.2012.11.008

• 人工智能 • 上一篇    下一篇

基于SVM的microRNA计算识别方法研究

孙萧寒,赵 维   

  1. 渭南师范学院数学与信息科学学院,陕西 渭南 714000
  • 收稿日期:2012-07-06 修回日期:1900-01-01 出版日期:2012-11-10 发布日期:2012-11-10

Computational Method for Identification of microRNA Based on SVM

SUN Xiao-han, ZHAO Wei   

  1. College of Mathematics and Information Science, Weinan Normal University, Weinan 714000, China
  • Received:2012-07-06 Revised:1900-01-01 Online:2012-11-10 Published:2012-11-10

摘要: microRNA(简称miRNA)是近年来发现的一类新的长约19-24nt的内源性非编码RNA,在动物和植物中发挥着重要而广泛的调控功能。它的发现主要有cDNA克隆预测和计算发现两条途径。由于cDNA克隆预测方法受miRNA表达的时间和组织特异性以及表达水平的影响,而计算发现可以弥补其不足,因此miRNA的计算方法受到了广泛的重视。支持向量机作为一种优异的机器学习方法,因其良好的主动学习能力可应用于miRNA的预测。本文利用支持向量机方法对线虫miRNA进行预测,结果表明,该方法具有良好的正确率和敏感性。

关键词: miRNA, SVM, 机器学习, 生物信息学

Abstract: microRNA (miRNA) is an endogenous non-coding RNA which is discovered in animals and plants. The length of miRNA is about 19-24nt. miRNA plays an important role and extensive functions in animals and plants. miRNA has two ways, cDNA clones predict and calculating found are the two ways for the miRNA discovery. cDNA clone predict method has some shortcomings, because the method can be influenced by the miRNA expression time, tissue-specific and expression levels. But the calculating found method can make up these shortcomings. So the miRNA calculating method has much attention. Support vector machine (SVM) as a kind of good machine learning method, is applied on the prediction of miRNAs for its good learning ability. The results of application of SVM on the identification of Caenorhabditis elegans miRNAs show SVM has good accuracy and sensitivity.

Key words: microRNA, SVM, machine learning, bioinformatics

中图分类号: