A hybrid metaheuristic algorithm for antimicrobial peptide toxicity prediction.

Area of Science:

• Computational Intelligence
• Optimization Algorithms
• Bioinformatics

Background:

• Developing novel algorithms is crucial for solving complex computational and real-world challenges.
• Existing optimization algorithms have limitations in exploration and exploitation balance.
• Hybridization of metaheuristic algorithms can potentially enhance performance.

Purpose of the Study:

• To introduce a novel hybrid metaheuristic algorithm, h-PSOGNDO.
• To combine the strengths of Particle Swarm Optimization (PSO) for exploitation and Generalized Normal Distribution Optimization (GNDO) for exploration.
• To evaluate the effectiveness of h-PSOGNDO on benchmark functions and a real-world application.

Main Methods:

• Developed a hybrid metaheuristic algorithm, h-PSOGNDO, integrating PSO and GNDO strategies.
• Tested h-PSOGNDO on twenty-eight IEEE CEC2017 and ten IEEE CEC2019 benchmark mathematical functions.
• Applied h-PSOGNDO to predict the toxicity of antimicrobial peptides.

Main Results:

• h-PSOGNDO demonstrated effective performance on both sets of benchmark test functions.
• The algorithm achieved highly competitive outcomes in predicting antimicrobial peptide toxicity.
• Statistical findings confirm the algorithm's efficacy for complex problem-solving.

Conclusions:

• The proposed h-PSOGNDO algorithm effectively leverages the advantages of both PSO and GNDO.
• h-PSOGNDO shows significant potential for addressing complex optimization problems.
• The algorithm's successful application in antimicrobial peptide toxicity prediction highlights its real-world applicability.