摘要
提出了高斯-经验模态分解(Gauss-EMD)算法,在固有模态函数筛选过程中,用高斯径向基函数插值极值包络线,并将该算法应用于DNA序列的相似性分析当中.这个方法可以简单直观地通过分析DNA序列的余量来分析序列的相似性,通过四个物种线粒体DNA序列实例验证了新算法相较于经典的经验模态分解算法更适用于DNA序列的局部相似性分析.
A Gauss empirical mode decomposition （Gauss-EMD） method is proposed and applied to the local similarity analysis of DNA sequences, using Gauss radial basis function to interpolate the extrema envelops during the intrinsic mode functions sifting process. The similarity among the DNA sequences is analyzed conveniently and intuitively just by their residues acquired by the Gauss-EMD method with the residues easy to be compared. By using the mitochondria of four different species, the suitability of the method is verified, and better performance is achieved for local similarity analysis of DNA sequences.
出处
《大连交通大学学报》
CAS
2013年第1期98-102,共5页
Journal of Dalian Jiaotong University
