DNA序列的分类模型

The Classified Model for DNA Sequences

本文针对 DNA序列分类这个实际问题 ,提出了相应的数学模型 .为了很好的体现 DNA序列的局部性和全局性的特征 ,我们给出了衡量分类方法优劣的标准 ,即在满足一定限制条件的情况下 ,是否能充分反映序列的各方面特性 .依据我们提出的判别标准 ,单一标准的分类是无法满足要求的 .我们的方法是侧重点不同的三种方法的综合集成 .这三种方法分别体现了序列中元素出现的概率 ,序列中元素出现的周期性 ,序列所带有的信息含量 .利用这个方法 ,完成了对未知类型的人工序列及自然序列的分类工作 .最后 ,对分类模型的优缺点进行了分析 。

Classifying the DNA sequences is a practice problem in biology. In this paper,a mathematics model is established for the classifying of DNA sequences.Since there are both locality and globality in the DNA sequences,we discuss the criterion about whether the classified method is good or not.That is whether the method bases on all properties that the DNA sequences have. So a classified method with a single standard is not enough for the problem.Here is a synthesis method on three different classified ways.The three ways base on varied property that DNA sequences have. The first is the frequency of occurrences of the elementin the DNAsequences.The second is the periodic property of the DNA sequences.The third is thatamount of information of the sequences. By using this method,we classify the nature sequences and artifical sequences. At last,we analyze the characteristic in this model and consider the generalization of this model