【作者】 李晟； 姜青山； 郭顺； 王备战；

【Author】 Li Sheng~1,Jiang Qingshan~(1,2),Guo Shun~1,and Wang Beizhan~1 1(Software School,Xiamen University,Xiamen,Fujian 361005) 2(Chengdu University,Chengdu 610106)

【机构】 厦门大学软件学院； 成都大学；

【摘要】 蛋白质序列作为生物序列数据一个重要组成部分,对其的分析研究已经成为生物信息学中的一个重要的研究方向和内容.通过对序列进行模式挖掘,可以对蛋白质序列或某一蛋白质家族序列进行研究,因此蛋白质序列的模式挖掘已经成为蛋白质序列研究中的一项重要任务.MBioPM是一种最新的生物序列模式挖掘算法,该算法通过引入模式划分概念,提高算法的效率,但该算法在效率方面仍存在不足,而且挖掘结果存在冗余性的问题.因此,提出一种优化算法BioPMMH,通过带有模式划分特点的Hash链表结构来优化算法中的搜索空间及策略,并在算法过程中对重复模式进行过滤.实验表明,算法BioPMMH能有效提高模式挖掘的效率,并解决结果的冗余性问题.更多还原

【Abstract】 As a part of biosequence,the analysis and research on protein sequences has already been an important topic and direction of Bioinformatics.Through motif mining,studies on a certain protein or a protein family then are able to be conducted;hence motif mining has been a major task in protein sequence study.MBioPM is one of the latest motif mining algorithms that is based on the idea of"motif-divided"to improve the efficiency.Concerning the efficiency and the result redundancy of MBioPM,an optimized BioPMMH algorithm is presented,which introduces the structure of motifdivided chained hash table to optimize the search strategy and space during the mining progress,and filters the repeated sub-motifs generated during the algorithm.Both theoretical analysis and experimental results have shown that the improved algorithm works better in the scale of running efficiency and solves the problem of redundancy.更多还原