Abstract: Based on the parallel optimization of gene computer clone software(SiClone) and alternative splicing analysis software(AltSplice), this paper puts forward a parallel optimization method. It divides needed large-scale sequence database into many segments according to some strategy, therefore each node is responsible for a smaller portion of the database, which eliminates disk I/O and heavy communication demands and much more fits for low-cost and efficent Linux clusters architecture.

Key words: parallel optimization, EST sequence database, segmentation deployment

摘要： 基于基因电脑克隆软件SiClone和可变剪接分析软件AltSplice的并行优化工作，提出一种基于大规模序列比对软件的并行优化方案。该方案对所要进行比对分析的大规模序列库按某种策略进行分割部署，使机群中每个节点处理整个序列库的一部分，可有效减少磁盘I/O和节点间通信开销，适合低价高效的Linux机群系统。

关键词: 并行优化, EST序列库, 分割部署

CLC Number: 

• TP311