The Fractal and Nonlinear Dynamical Properties of Signal Peptides

We analyzed the Hurst exponent for the amino acid sequences of proteins and secretory signal peptides using the aggregated variance method. Then the diversification entropy of the sequences was calculated. The numeralization of the amino acid sequences are according with the three amino acid indices: EIIP, polarity and accessible surface area. The results show that when an amino acid sequence coding with EIIP or polarity indices has both signal peptides and protein parts, the diversification entropy of signal peptide is less than the diversification entropy of protein. But if coding with accessible surface area, there is no evidence different. We take the two parts´ Hurst exponents as a vector characterizing the sequence, and it can be found that the diagrams of the sequences having signal peptides are different from the diagrams of the sequences having non-signal peptides