Chaotic step length artificial bee colony algorithms for protein structure prediction
Akash Saxena, Shalini Shekhawat, Ajay Sharma, Harish Sharma, Rajesh Kumar
Abstract
Abstract Protein Structure Prediction (PSP) is an important problem of bio informatics. It has a paramount importance due to the fact that accurately predicted structure can help scientist and doctors to fight with the chronic diseases. Recently some studies reported efficacy of the bio inspired algorithms to solve these problems reported in the literature. It has also been a known fact now that Fitness Landscape Analysis (FLA) of Artificial Bee Colony Algorithm (ABC) is suitable for these problems. Inspired with these facts, authors of this paper present new variants of ABC as Chaotic Step Length driven ABC algorithms named as Enhanced Chaotic Artificial Bee Colony Algorithms (ECABC). In these variants, 10 different Chaotic Step Lengths (CSLs) are embedded in the position update phase of onlooker bee phase. 10 different variants based on CSL are proposed and evaluated on artificial and real protein structure benchmarks. Different statistical analyses reveal that the performances of ECABCs are superior to the parent algorithm. These algorithms may become a suitable choice to solve protein structure problem.