A comparison between artificial bee colony and particle swarm optimization algorithms for protein structure prediction problem

Abstract

Protein Structure Prediction (PSP) is a well known problem for Bioinformatics scientists. It was considered as a NP-hard problem. Swarm Intelligence is a branch of evolutionary algorithm, is commonly used for PSP problem. The Artificial Bees Colony (ABC) optimization algorithm is inspired from the honey bees food foraging behavior and the Particle Swarm Optimization (PSO) algorithm which also simulate the process of the birds' foraging behavior are both used to solve the PSP problem. This paper investigates the performance of the two algorithms when being applied on an experimental short sequence protein called Met-enkaphlin in order to predict its 3D structure. The results illustrates clearly the power of the PSO search strategy and outperforms the ABC in terms of Time, Avg.NFE and success rate values by 70%, 73%, 3.6% respectively. However, the ABC results were more stable than the PSO in terms of Std.dev values, by 74%. © Springer-Verlag 2013.