The Double-Funnel Energy Landscape of an Off-Lattice Model Protein: A Knowledge-Based Evolution Algorithm Approach

摘要

We apply an off-lattice minimal energy model of proteins and the Knowledge-based Evolution Algorithm (KEA)to study the double-funnel landscape of protein folding problem. The testing off-lattice minimal energy model is composed of a polypeptide chain, and has the Ramachandran angles as its degrees of freedom. The force field of this model is based on hydrogen bonds and the anisotropic hydrophobicity forces. The Knowledge-based Evolution Algorithm tries to simulate the process of the knowledge development. The evolutionary knowledge database will help to direct searching processes and to reach the global minimum faster.Our results show that the KEA can improve the searching efficiency.In this report, we will illustrate an off-lattice minimal energy model and how to apply KEA to this model to study its double-funnel energy landscape. The phase transition of -helix and -hairpin structures and the evolution of guiding function are reported.