Study On The Characteristics Of Biological Sequences Based On Nonlinear Theory Methods

To explore biological information from biological data is the long-range objective of Bioinformatics. There are complicated relations between these data, which also contain abundant biological knowledge and biological rules. In this paper, we mainly study the property of biological sequences using the nonlinear theory methods.After introducing the background of Bioinformatics, we first study the characteristic of amino acid sequences, which is one precondition for analyzing the relationship between amino acid sequences and protein structures. Based on the former researches, the nonlinear prediction method is introduced firstly, and used for analyzing two groups of protein sequences. It is found that protein sequences are provided with nonlinear deterministic property. In order to study this property more clearly, the same method is used to study random sequences and chaotic sequences. By comparing the results of protein sequences with random sequences and chaotic sequences, it is found that protein sequences are nonrandom, provided with obvious nonlinear property, which may be chaos.Calculating the largest Lyapunov exponents is the most direct way to judging if one system is provided with chaos. The main innovation of this paper is to calculate the largest Lyapunov exponents for protein sequences. Firstly, protein sequences are translated into time series by using the chaos game representation method. Then the largest lyapunov exponents are computed from the time series by using Wolf's algorithm. Under the selected delay time and embedding dimension, the largest lyapunov exponents of two groups of protein sequences are all greater than zero. By Chaos theory, the Sufficient Conditions of a system existing chaos are the largest lyapunov exponents of this system are greater than zero. Thus the conclusion is reached: protein sequences have chaotic characteristics.For DNA analysis, we first cite a method describing as DNA sequences'genomic signature. Then apply this method to the selected alien invasive plants'DNA sequences. It is an attempt for applying theory to practice. It will be helpful to alien invasive plants'species identification, species evolution and classification.