Investigation of the three-dimensional lattice HP protein folding model using a genetic algorithm

Abstract

An approach to the hydrophobic-polar (HP) protein folding model was developed using a genetic algorithm (GA) to find the optimal structures on a 3D cubic lattice. A modification was introduced to the scoring system of the original model to improve the model's capacity to generate more natural-like structures. The modification was based on the assumption that it may be preferable for a hydrophobic monomer to have a polar neighbor than to be in direct contact with the polar solvent. The compactness and the segregation criteria were used to compare structures created by the original HP model and by the modified one. An islands' algorithm, a new selection scheme and multiple-points crossover were used to improve the performance of the algorithm. Ten sequences, seven with length 27 and three with length 64 were analyzed. Our results suggest that the modified model has a greater tendency to form globular structures. This might be preferable, since the original HP model does not take into account the positioning of long polar segments. The algorithm was implemented in the form of a program with a graphical user interface that might have a didactical potential in the study of GA and on the understanding of hydrophobic core formation.

Protein-folding simulations of the hydrophobic-hydrophilic model by combining pull moves with energy landscape paving

“…The tested instances include ten 2D benchmark sequences with length ranging from 20 to 100 [12] and ten 3D ones with length 64 [3]. These benchmark instances, in part, have been used extensively in the literature [3][4][5][6][7][8][9][10][11][12][13][14][15][16][17][18][19][20][21]. A complete listing of the 2D and 3D sequences can be found in Table I and II, respectively.…”

“…Therefore, ELP-pull moves explores the conformation surfaces more efficiently than MC, GA, SISPER, GAOSS, MC-pull moves and GDMC-pull moves. For ten 3D sequences with 64-mers [3], we list the results of the ELP-pull moves algorithm in Table IV, in comparison with those by Monte Carlo (MC) [3], genetic algorithm (GA) [3], improved genetic algorithm [IGA] with a new selection scheme and multiple-points crossover [4], guided genetic algorithm (GGA) [5], and particle swarm optimization (PSO) [6]. From Table IV one can see that ELP-pull moves finds new lower free energies than these five methods for all ten 3D sequences.…”

“…In order to search for the minimum energy conformations for HP model, people have presented many heuristic algorithms, including genetic algorithm (GA) [3], and its variations (improved genetic algorithm (IGA) with a new selection scheme and multiple-points crossover [4] and guided genetic algorithm (GGA) [5]), particle swarm optimization (PSO) [6], evolutionary Monte Carlo (EMC) [7], varieties of Monte Carlo (MC) [8], the hybrid of GA and tabu search (GTS) [9], elastic net algorithm with local search (ENLS) [10,11], gradient-directed Monte Carlo (GDMC) [12], GA based on optimal secondary structures (GAOSS) [13], and hybrid Taguchi genetic algorithm (HTGA) [14] that combines genetic algorithm, Taguchi method, and PSO.…”

Protein Folding in Hydrophobic-Polar Lattice Model: A Flexible Ant-Colony Optimization Approach

“…Various computational methods have been applied in PFPs with the HP square lattice model, such as genetic algorithms (GAs) [25][26][27][28][29], Monte Carlo (MC) algorithms [30][31][32][33][34][35], Immune system (IM) algorithms [36,37] and ant colony optimization (ACO) [38][39][40][41] etc. Since the first ant algorithm -ant system (AS) was proposed by Dorigo [42], successors such as the elitist ant system (EAS), max-min ant system (MMAS), and ant colony system (ACS) [43,44] etc.…”

Protein structure prediction with the 3D-HP side-chain model using a master–slave parallel genetic algorithm

“…This fact has motivated the development of many metaheuristic approaches for dealing with the problem. In this scenery, evolutionary computation methods and, in special, Genetic Algorithms (GA) have been proved not only adequate, but very efficient [11,20,26].…”