Emerging traits in the application of an evolutionary algorithm to a scalable bioinformatics problem

In this work we present an empirical study on the application of a genetic algorithm featuring rank based variation operators (named Rank GA) to one scalable problem in the area of bioinformatics (Protein Folding in a 2D square lattice) trying to discover the main traits that emerge as the scale of the problem grows. This study has a double intention: 1) to develop a robust easy-to-understand evolutionary algorithm for solving the protein folding problem and 2) to grasp some more general and theoretical intuition from this kind of difficult problems that often are being approached through evolutionary algorithms. We show how the use of a squared 2D lattice in the model can influence the outcome of the Rank GA and also how this algorithm compares in performance with previously published results. The Rank GA seems to perform better than other EAs as the problem size scales up.