An efficient hybrid Taguchi-genetic algorithm for protein folding simulation
Introduction
The prediction of protein structure from its amino-acid sequence is one of the most prominent problems in computational biology. A protein’s function depends mainly on its tertiary structure, which in turn depends on its primary structure. Mistakes in the folding process create proteins with abnormal shapes, which are the causes of diseases such as cystic fibrosis, Alzheimer’s, and mad cow. If we could predict the tertiary structures of proteins from their sequences, we would be able to treat these diseases better. The knowledge of protein tertiary structures also has other applications, such as in the structure-based drug design field (Bui & Sundarraj, 2005).
Currently, protein structures are primarily determined by techniques such as MRI (magnetic resonance imaging) and X-ray crystallography, which are expensive in terms of equipment, computation, and time. Additionally, these techniques require isolation, purification, and crystallization of the target protein. Computational approaches to protein structure prediction are therefore very attractive. The difficulty in solving protein structure prediction problems stems from two major sources: (1) finding good measures for the quality of candidate structures, and (2) given such measures, determining optimal or close-to-optimal structures for a given amino-acid sequence (Krasnogor, Hart, Smith, & Pelta, 1999).
Recently, many researchers (Krasnogor et al., 1998, Krasnogor et al., 1999, Patton et al., 1995, Pedersen and Moukt, 1997) have used evolutionary algorithms, such as the genetic algorithms (GA), for solving the protein folding problem. Genetic algorithms are stochastic search techniques based on the mechanism of natural selection, which requires information to search effectively in a large or poorly understood search space. The effectiveness of crossover and mutation is weakened in the protein folding problem (Krasnogor et al., 1998, Krasnogor et al., 1999), since by increasing the compact folded structure, the failure of the crossover operation increases due to collisions. Further, in sequences of mutation, there will be often invalid conformations due to collisions within compact conformation. Therefore, some researchers (Bui and Sundarraj, 2005, Jiang et al., 2003, König and Dandekar, 1999, Takahashi et al., 1999) have proposed various hybrid methods to improve GA. The above-mentioned improved GA methods were mainly aimed at the crossover and mutation operations.
Recently, the Taguchi method is a robust design approach. It uses many ideas from statistical experimental design for evaluating and implementing improvements in products, processes, and equipment. The fundamental principle is to improve the quality of a product by minimizing the effect of the causes of variation without eliminating the causes. The Taguchi method is suitable for a wide range of applications (Kaytakoğlu and Akyalçın, 2007, Liu et al., 2007, Wang and Huang, 2008), including the following practices: quality engineering, experimental design, business data analysis, management by total results, pattern recognition, and so on. The Taguchi method is a series of approaches that predicts and prevents troubles or problems that might occur in the market after a product is sold and used by a customer under various environment and real-life conditions for the duration of the product life.
In this paper, we focus on the 2D hydrophobic-polar (HP) lattice model. An efficient hybrid Taguchi-genetic algorithm (HTGA) is proposed for solving the protein folding problem in the 2D HP model. The proposed HTGA is mainly aimed to improve the crossover and mutation operators and enhance exploitation capability. In order to improve the crossover operation, we use the Taguchi method to select the better genes. In the mutation operation, we employ the merits of PSO to improve the mutation mechanism. Lin et al., 2008, Lin and Hong, 2007 used the particle swarm optimization to improve the mutation mechanism. Simulation results show that our method has a better performance than those of existing methods in protein folding problem.
The remainder of this paper is structured as follows: Section 2 gives the preliminaries and the formal definition of the protein folding problem in the 2D HP lattice model. Section 3 describes our approach in detail. The proposed hybrid Taguchi-genetic algorithm combining the traditional genetic algorithm, the Taguchi method, and particle swarm optimization is presented. The experimental results obtained by our method and by other methods are compared in Section 4. Finally, the conclusion is given in the last section.
Section snippets
Preliminaries
In this section, we briefly present the 2D HP protein folding problem and its free energy calculation.
Methods
In this section, we review the Taguchi method and particle swarm optimization. An efficient hybrid Taguchi-genetic algorithm is also presented.
Simulation results
In this section, we compared our method with the traditional genetic algorithm (Unger & Moult, 1993), the ant colony algorithm (Shmygelska, Anguirre-Hernandez, & Hoos, 2002), Monte Carlo (Liang & Wong, 2001), and the tabu search with the genetic algorithm (Jiang et al., 2003). In Table 4, the 8 chosen HP instances are standard benchmarks used to test the searching ability of the algorithms. The free energy is the optimal or best-known energy value. indicates i repetitions of the
Conclusion
This paper proposed an efficient hybrid Taguchi-genetic algorithm for solving protein structure prediction problem. The proposed hybrid algorithm combines the genetic algorithm, Taguchi method, Particle Swarm Optimization, and a local search of the protein folding pathway. The Taguchi method was used obtain the optimum offspring, while the particle swarm optimization was used to improve mutation operation of the genetic algorithm. In mutation operation based on particle swarm optimization, the
References (27)
- et al.
Optimal HP configurations of proteins by combining local search with elastic net algorithm
Journal of Biochemical and Biophysical Methods
(2007) - et al.
Optimization of parametric performance of a PEMFC
International Journal of Hydrogen Energy
(2007) - et al.
The design of neuro-fuzzy networks using particle swarm optimization and recursive singular value decomposition
Neurocomputing
(2007) - et al.
Protein folding simulations with genetic algorithms and a Detailed Molecular Description
Journal of Molecular Biology
(1997) - et al.
Genetic algorithms for protein folding simulations
Journal of Molecular Biology
(1993) - et al.
Optimizing back-propagation networks via a calibrated heuristic algorithm with an orthogonal array
Expert Systems with Applications
(2008) - Bui, T. N., & Sundarraj, G. (2005). An efficient genetic algorithm for predicting protein tertiary structures in the 2D...
- et al.
Genetic fuzzy systems evolutionary tuning and learning of fuzzy knowledge bases
(2001) Theory for the folding and stability of globular proteins
Biochemistry
(1985)- Eberhart, R., & Kennedy, J. (1995a). A new optimizer using particle swarm theory. In Proceedings of the sixth...
Particle swarm optimization
IEEE International Conference on Neural Networks
Protein folding simulations for the hydrophobic–hydrophilic model by combining tabu search with genetic algorithms
Journal of Chemical Physics
Refined genetic algorithm simulation to model Proteins
Journal of Molecular Modeling
Cited by (27)
Protein folding in 3D lattice HP model using a combining cuckoo search with the Hill-Climbing algorithms
2022, Applied Soft ComputingMultimodal Memetic Framework for low-resolution protein structure prediction
2020, Swarm and Evolutionary ComputationPopulation-based sampling and fragment-based de novo protein structure prediction
2018, Encyclopedia of Bioinformatics and Computational Biology: ABC of BioinformaticsAPL: An angle probability list to improve knowledge-based metaheuristics for the three-dimensional protein structure prediction
2015, Computational Biology and ChemistryCitation Excerpt :In Kondov (2013), a distributed parallel Particle Swarm Optimization algorithm was developed for protein structure prediction problem. Lin and Hsieh (2009) presents a hybrid PSO/GA algorithm to search for the native structure of a protein molecule in a hydrophobic-hydrophilic lattice model representation. Despite the advances, metaheuristics still have to deal with the challenge of vast conformational search spaces caused by the different combination of amino acid residues.
A novel state space representation for the solution of 2D-HP protein folding problem using reinforcement learning methods
2015, Applied Soft Computing JournalCitation Excerpt :Thus, in this study the reinforcement learning methods are used for the solution of the protein-folding problem in two dimensional lattice model. There exist many studies [4–9] in literature that proposed different methods for the solution of this problem, but the use of reinforcement learning methods are quite new. In [10–13], authors used the Q-learning algorithm to solve the protein folding problem in two dimensional hydrophobic-polar (2D-HP) model.
Improved Takagi-Sugeno fuzzy model-based control of flexible joint robot via Hybrid-Taguchi genetic algorithm
2014, Engineering Applications of Artificial IntelligenceCitation Excerpt :There are two major tools in the Taguchi method: orthogonal array (OA) and the signal to noise (S/N) ratio (Tsai et al., 2004). The detailed flowchart of this method has been shown in Fig. 1 (Lin and Hsieh, 2009). OAs have different level types but in this paper two-level orthogonal arrays are used.