首页> 外文会议>International Conference on Network and Information Systems for Computers >Distributed Parallel Needleman-Wunsch Algorithm on Heterogeneous Cluster System
【24h】

Distributed Parallel Needleman-Wunsch Algorithm on Heterogeneous Cluster System

机译:在异构群体系统上分布式并行针头武警算法

获取原文

摘要

At first, the needleman-wunsch algorithm that relies the optimal matching result is regarded as the basic algorithm to solve the problem of sequence alignment. Nevertheless, as the result of sequence database capacity index increased year by year and the algorithm of time and space complexity are higher, O (m*n). Facing these problems, researchers need to reduce the sequence alignment algorithm becomes more important for time and space complexity basic algorithm to solve the problem of sequence alignment. Therefore, this paperuses the characteristics of heterogeneous cluster , basing on the theory of divisible load, using the communication strategy of LIFO, simplex communication model, the design bases on dual sequence global alignment parallel algorithm of Needleman-Wunsch, the algorithm puts forward the strategy of partitioning score data matrix , to determine the optimal number of iterations and assigns to the child nodes of sub-sequence length ability to making full use of each node and performs tasks in a row, at the same time puts forward the strategy of backtracking, reduces the algorithm operation time, makes integral algorithm optimal.
机译:首先,依赖最佳匹配结果的针对Wunsch算法被认为是解决序列对齐问题的基本算法。然而,由于序列数据库容量指数的结果逐年增加,时间和空间复杂性算法较高,o(m * n)。面对这些问题,研究人员需要减少序列对齐算法对时间和空间复杂性基本算法变得更加重要,以解决序列对齐问题。因此,本文的特点是异构集群的特点,基于可分离的负载理论,利用LIFO的通信策略,单纯X通信模型,设计基础对双序的双序全局对准并行算法,算法提出了策略分区评分数据矩阵,以确定最佳迭代次数并分配给子序列长度的子节点,以充分利用每个节点并连续执行任务,同时提出了回溯策略,减少算法操作时间,使得整体算法最佳。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号