CAJ | 학술논문

基于剪接位点竞争机制,剪接位点对分成竞争性剪接位点对和非竞争性剪接位点对.并且竞争性和非竞争性剪接位点对的分类是一个很重要的工作.结合位置权重矩阵、离散量和支持向量机,提出了预测竞争性和非竞争性剪接位点对的新方法.独立检验集中90%以上的剪接位点对能被正确地分类成竞争性和非竞争性剪接位点对.此预测成功率高于其它方法.
기우전접위점경쟁궤제,전접위점대분성경쟁성전접위점대화비경쟁성전접위점대.병차경쟁성화비경쟁성전접위점대적분류시일개흔중요적공작.결합위치권중구진、리산량화지지향량궤,제출료예측경쟁성화비경쟁성전접위점대적신방법.독립검험집중90%이상적전접위점대능피정학지분류성경쟁성화비경쟁성전접위점대.차예측성공솔고우기타방법.
Based on the mechanism of splice site competition, splice site pairs are divided into competitive splice site pairs and non-competitive splice site pairs, and the classification of competitive and non-competitive splice site pairs is an important task. The support vector machine method which combines with position weight matrix and increment of diversity is proposed for predicting competitive and non-competitive splice site pairs. The predictive results for the independent testing sets show that more than 90% of the splice site pairs are correctly classified into competitive and non-competitive. While compared with recent method, the predictive capability and reliability are improved by using our method.