2D HP modeli için dalgalanan benzetilmiş tavlama ile birleşmiş genetik algoritma

Protein folding is the process of a linear chain of amino acids folding into a functional 3D native structure. The problem of predicting this 3D native structure given only the amino acids of a protein is one of the most challenging problems in computational biology and is still rigorously investigated. The Hydrophobic-Polar model is a simplified model for the protein folding process. This model is based on the assumption that hydrophobicity is the dominant force that drives the protein folding process.In this thesis we propose a meta-heuristic algorithm (GAOSA) that combines the well known Genetic Algorithm approach with Oscillating Simulated Annealing to address the protein folding problem in the simplified 2D Hydrophobic-Polar Model. Our approach makes use of the pull move neighborhood for the mutation operator, a brute force 1-point crossover operator, a memory component borrowed from Tabu Search and a problem specific diversification phase. We also provide some insights about the implications of the Hydrophobic-Polar Model and how these implications can be utilized in an algorithm.