基于支持向量机的供体剪接位点识别

Recognition of donor splice sites based on support vector machine

剪接位点识别是基因识别中的关键环节。本文对待测样本采用0/1编码,以表征各位置上的碱基,并结合碱基二联体出现的频次,最后采用支持向量机(SVM)进行分类决策。HS3D数据集上的仿真结果显示,本方法获得的预测精度为92.84%。

splicing site recognition is the key link in gene recognition. In this paper, the test samples are encoded with 0 / 1 encoding to represent the bases at each position, and the occurrence frequency of the base dimer is combined. Finally, support vector machine ( SVM) is adopted to make classification decisions. The simulation results on HS3D data set show that the prediction accuracy obtained by this method is 92. 84%.