3D Protein Structure Prediction with Genetic Tabu Search Algorithm in Off-Lattice AB Model

摘要

A novel hybrid algorithm GATS (Genetic Tabu Search), which combines Genetic Algorithm (GA) and Tabu Search (TS) Algorithm, is presented for dealing with multi-extremum and multi-parameter problem based on AB off-lattice model in the predicting 3D protein folding structure. Tabu Search is introduced into the crossover and mutation operators in the Genetic Algorithm to improve the local search ability, and a variable population size is designed to keep the diversity of population. Experimental results show that the lowest energies computed by our GATS algorithm are better than those obtained by previous methods, and the lowest-energy conformation of a given protein sequence obtained by GATS forms a single hydrophobic core as observed in real proteins. Compared with previous approaches, GATS algorithm has higher efficiency and can predict 3D protein structure more effectively.