Motif识别算法简介及软件性能的研究.docVIP

Motif识别算法简介及软件性能的研究 　　摘 要：Motif在转录和后转录水平的基因表达调控中起着重要的作用。目前，识别Motif的算法和相应的软件已有不少，但是却鲜有对各种算法及软件性能共同评测的研究和报告。介绍了算法的分类以及三种常见的Motif识别算法Wordup，MM和Gibbs采样，并对AlignACE，MEME，MotifSampler，Weeder等13种Motif寻找软件进行性能比较分析。通过生物学意义的研究和性能比较结果可以得出：由于唯有Weeder算法考虑了Motif保守核心位置，因而它在各种软件中识别效果较好；大部分算法只考虑简单而且短的Motif，所以各种软件对酵母菌这种单细胞生物的Motif识别性能比多细胞生物要高。 　　关键词：Motif； Wordup； MM； Gibbs采样 　　中图法分类号：TP301．6 文献标识码：A 文章编号：1001-3695(2006)10-0066-04 　　Introduction of Algorithms and Performance Research of 　　Softwares for Motif Discovery 　　ZHU Ji1,2, YANG Hua1,2, NIU Bei??fang1,2, LANG Xian??yu1,2, LU Zhong??hua??1, CHI Xue??bin??1 　　(1.Supercomputing Center, Computer Network Information Center, Chinese Academy of Sciences, Beijing 100080, China; 2.Graduate School, Chinese Academy of Sciences, Beijing 100049, China) 　　Abstract:Motif plays a key role in the gene??expression regulating on both transcriptional and post??transcriptional levels. Nowadays there are several algorithms and softwares on detecting Motif, but, however, there is few papers on comparing the performance of these algorithms and softwares. This paper comes up with this background to introduce the classification of the algorithms in general and three common algorithms: Wordup, MM, Gibbs sampling??in details. And a performance comparison is made on the thirteen softwares for Motif detecting such as AlignACE, MEME, MotifSampler, Weeder, etc. Based on the biological research and the performance report, this paper ends with a conclusion that Weeder is the most effective one of these softwares, for it is the only algorithm that takes account of the conserved core positions of Motifs; Most algorithms only consider simple and short Motifs, so their Motif detecting performance on monadic yeast is significantly higher than on metazoans. 　　Key words:Motif; Wordup; MM(Mixture Model); Gibbs Sampling 　　 　　基因非编码区的一个主要研究方向是对Motif的研究。因为转录和后转录水平，其基因的表达在很大程度上受到一些Motif的控制。它们本质上是一些比较短的DNA序列，这些序列一般均处在受调控基因的上游区域，转录因子可识别这些Motif并与之结合，从而调节DNA的代谢和转录；或