多重序列比对Alignment的信息度量准则

Information Measure Criteria of Multiple Sequences Alignment(MSA)

多重序列的比对(Alignment)的核心问题是对多重DNA(或RNA、蛋白质)序列,寻找它们的相似部分或稳定区域,但如何定义多重序列的相似度(或罚分函数)则是解决多重序列比对的前提。首先对多重序列相似度的合理性进行了讨论,给出了一组合理的相似度所必需具备的条件;再利用Shannon熵的特点,给出一个满足这些合理性条件的多重序列比对的优化准则。

The key of the alignment problem for the multiple biology sequences(DNA, RNA, or Pritein) is finding their similar part or stable region. So, the definition of the similarity measure or the cost function is a presupposition for the MSA. In this paper, we first give some rationality conditions of the similarity, then we give an optimization measure criterion using the Shannon entropy. This optimization measure criterion is a satisfaction to these similarity conditions.