Study on the chaotic characteristics of protein sequences

In this paper, the nonlinear prediction technique is used to study chaotic sequences, random sequences and protein sequences. The chaotic sequences are from the specific logistic system, and protein sequences are selected from Class, Architecture, Topology, Homology (CATH) database. From the results of this method, it is found that protein sequences may be chaos. In order to verify this hypothesis, Protein sequences are translated into time series by using electron-ion interaction pseudopotential (EIIP) values. Then the largest Lyapunov exponents are computed for the time series, which are greater than zero. Thus we reach the conclusion: protein sequences have chaotic characteristics.