Improved Particle Swarm Optimization Algorithm for 2D Protein Folding Prediction

One of the main problems of protein folding prediction is optimization computation. On basis of toy model, this paper proposes an improved particle swarm optimization algorithm for protein folding prediction. The algorithm introduces a new architecture that is characterized by balancing exploration and exploitation capability of particle swarm optimization algorithm. In the architecture, the population in each generation consists of three parts: an elitist part, an exploitative part, and an explorative part. In the meantime, it makes protein folding prediction more effective with the global search and local search ability in the improved algorithm. Furthermore, we have applied the improved particle swarm optimization algorithm to sequences with up to 55 monomers within toy model and the experimental results are compared with the global energy minimum reported in some literatures. It demonstrates that the proposed algorithm is effective to search for the native state of proteins with the lowest free energy.