摘要
在基因序列图形表达模型研究的基础上,提出了一种新的非退化的基因图形三维表示方法。该表达方法不仅避免了图形的重叠和交叉,同时还保留了序列的生物学特征。利用该表达方法对H5N1病毒基因序列进行数字特征的提取并引入基于多维PFS判别函数进行模糊聚类分析应用。在聚类分析过程中直接利用数字特征矩阵作为分析数据,分析结果表明:利用文中所给图形表达建立基因序列数字特征矩阵进行的聚类分析具有一定的合理性。
This article proposed a new 3D graphic representation with non-degeneration on the basis of studying gene sequences representation model. This representation can avoid the overlap or cross without losing biological information. This method was used to abstract character of I-ISN1 virus gene and propose a clustering analysis based on multi-dimension Pseudo F-Statistios (PFS) parameter. In the procedure of clustering, the numerical characterization matrix was used as analytic data source. The result shows that it is rational to make clustering analysis on the gene character abstract from the new graphic representation.
出处
《计算机应用》
CSCD
北大核心
2007年第9期2330-2333,共4页
journal of Computer Applications
基金
湖南省自然科学重点基金资助项目(06JJ4076)
湖南省财政厅项目基金资助项目([2005]90)
关键词
基因序列
图形表达
聚类分析
伪F统计量
基因数字特征
gene sequences
graphic representation
e.lustering analysis
Pseudo F-Statistios
gene numerical characterization