Identification Of Tandem Repeats Using Spectral Analysis

Tandem repeat identification is a significant, challenging issue on Bioinformatics, it has very important significance for discovering the function and heredity of some specific sequences in gene sequences. With the accomplishment of gene sequencing works, a large number of gene sequences was identified and stored in gene databases, and need to be processed and analyzed. It has been discovered that a lot of tandem repeat identification methods have been proposed. These methods were divided into two classes, the first kind of methods is based on string matching method, and the other is based on digital signal processing. This paper will research on the second class, after analyzing domestic and foreign tandem repeat identification methods in this field, find that these methods still have some drawbacks in calculation amount and identification accuracy. This paper is aim at the large amount of calculation and low accuracy in existing method.Firstly, introduces some concepts in tandem repeat and some foundation theories in biology. Deeply researches tandem repeat identification method status at home and abroad, and elaborates advantages and disadvantages of some representative identification methods.Secondly, analyzes the advantages and disadvantages of tandem repeat identification methods based on Discrete Fourier Transform, and improves these existed drawbacks. Presents a gene sequence digital presentation method based on nucleotide EIIP, employs the EIIP of each nucleotide as the numerical representation for DNA sequence. A gene sequence can be expressed as a unique digital sequence, reduced the calculation amount for compute the spectrum.Thirdly, deeply analyzes the PSE method, for the drawbacks of this method, presents a tandem repeat identification method based on AR model. Resolves the large computation, inaccurate of order estimation and spectral splitting problem may occur in PSE method.Finally, verifies and analyzes methods presented in the paper by experiment. Compare and analyze the experimental results with the existing method and the information labeled in GenBank, shows the correctness and affectivity of the proposed methods.