Solving the Protein Structure Prediction Problem Through a Multiobjective Genetic Algorithm
Richard O. Day, Gary B. Lamont, Ruth Pachter, Jesse B. Zydallis · 2002
1 The Protein Structure Prediction problem, which involves correctly predicting the geometrical conformation of a fully folded protein, is extremely difficult to solve and there currently is no “best ” method of generating solutions. This paper focuses on an energy minimization technique and the use of a multiobjective genetic algorithm, the multiobjective fast messy genetic algorithm (fmGA) to obtain solutions to this problem. We extend the fmGA to generate solutions to the PSP problem as a multiobjective problem using the CHARMm energy function. Further, the results of the multiobjective fmGA formulation compare very favorably to our previous results from the single objective fmGA formulation.