PROTEIN FOLDING PADA LATTICE HP MODEL MENGGUNAKAN DIFFERENTIAL EVOLUTION

• Main Author: Ramadhian Aribowo, Beryl
• Format: Theses
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/49071
• Institution: Institut Teknologi Bandung

The goal of protein folding is to predict the protein native structure given a sequence of amino acid. Protein folding is viable to be modelled as a mathematical optimization problem, for instance by transforming the structure and properties of an amino acid sequence into lattice model. Within this research, differential evolution algorithm is employed to optimize a protein objective function under single target global optimum assumption. The combination of clustering and differential evolution is adopted to optimize the protein objective function under multimodal function mapping problem. The optimization simulation has successfully obtained the global optimum points of amino acid sequence with length 20 (UM20) on both single target global optimum problem and multimodal problem. The protein conformations are also analyzed specifically on multimodal problem formulation. On multimodal problem formulation, it has been discovered that if there exist several different global optimum points, the final protein conformations generated are equivalent, which signifies that an amino acid sequence only yields one stable protein conformation.