Smith-Waterman算法的若干优化及并行实现  

Parallelization and Optimization of Smith-Waterman

作　　者：周澄[1] 郁松年[1] 

机构地区：[1]上海大学计算机工程与科学学院,上海200072

出　　处：《计算机工程与应用》2003年第23期89-91,共3页Computer Engineering and Applications

基　　金：上海市科委重点基金项目支持(编号:00JC14052)

摘　　要：Smith-Waterman算法是目前被使用最广泛的序列相似性比较算法之一,它适用于寻找局部相似序列对。该算法精确度较高,一直沿用到现在。目前,使Smith-Waterman算法提速,寻找该算法的优化方法,是世界各地的科学家们正花费大量心血研究的课题。该文从算法并行化着手,充分利用近期蓬勃发展的高性能计算机系统,提出了若干Smith-Waterman算法的优化思想,并在cluster机上实现。Smith-Waterman algorithm is one of the most widely used algorithms ,which finds local alignments between sequences in genome.It has been used till now for its high precision in comparison.To improve and optimize the algorithm,researches never stop world-widely.This article focuses on parallelization of this algorithm,and implement it on clustering computers,to make good use of the high-performance and massively-parallel computing technology appeared these years.

关 键 词：生物信息学 序列比对算法 算法并行化 集群机 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]