Genetic Algorithm Variants in Predicting Protein Structure

Proteins are the machinery of life and common to all organisms. In Protein Structure Prediction (PSP) the tertiary structure of a protein is predicted by using its primary structure information. It can help in the design of new drugs and medicines. As PSP problem has been proved to be an NP-hard problem we go for meta-heuristic techniques to solve it. In this paper, we have taken six variants of Genetic Algorithms (GA), applied them in predicting protein structure and compared their performances. As GA has several genetic operators, such as, selection, crossover and mutation, we can modify them to improve the overall performance. On the basis of selection we have considered three variants: GA1 uses rank selection method, GA2 uses elitist selection method and GA3 uses tournament selection method. All these variants are implemented taking two types of crossovers, such as, single point crossover and double point crossover. In this way, six variants have been implemented. It is observed that GA2 with two point crossover outperforms other variants in minimizing energy.