Inter-Algorithm Multiobjective Cooperation for Phylogenetic Reconstruction on Amino Acid Data.

Area of Science:

• Bioinformatics
• Computational Biology
• Evolutionary Computation

Background:

• Evolutionary algorithms (EAs) are increasingly used for complex optimization.
• Cooperative EAs enhance search capabilities but require advanced designs for real-world problems.
• Phylogenetic history reconstruction from amino acid data is a challenging bioinformatics problem.

Purpose of the Study:

• To develop and evaluate a multiobjective cooperative evolutionary algorithm for phylogenetic reconstruction.
• To address the growing complexity of optimization problems in bioinformatics.
• To improve solution quality in evolutionary computation through novel cooperative strategies.

Main Methods:

• A multiobjective cooperation framework integrating three distinct EA trends: Nondominated Sorting Genetic Algorithm II, Indicator-Based Evolutionary Algorithm, and Multiobjective Evolutionary Algorithm based on Decomposition.
• An Elite island component for managing algorithm migrations and performance feedback.
• Dynamic reallocation of computational resources to the most effective algorithms during execution.

Main Results:

• The proposed cooperative approach significantly enhances performance in reconstructing phylogenetic histories from amino acid data.
• Demonstrated superior results compared to stand-alone algorithms, standard island models, and state-of-the-art methods.
• Effectively handles complex optimization tasks in bioinformatics.

Conclusions:

• Multiobjective cooperation offers a powerful strategy for advancing evolutionary algorithms in complex problem domains.
• The developed method provides a robust solution for phylogenetic reconstruction.
• Performance-driven cooperative designs are crucial for tackling sophisticated real-world optimization challenges.