PSO-GA cooperative optimization algorithm based protein structure prediction

Low homology protein tertiary structure prediction with huge number of local minimums is investigated. A novel PSO-GA cooperative optimization algorithm, which combines particle swarm optimization with genetic algorithm, is proposed by sharing the overall best solution. Simultaneously, the optimization is accelerated by running the two algorithms parallelly. Simulation results show that the precision of low homology protein tertiary structure prediction based on the proposed PSO-GA algorithm reaches 0.6 nm, much close to the international standard and much better than that by PSO or GA alone, with much better convergence rate.