Applying a multiobjective metaheuristic inspired by honey bees to phylogenetic inference

Area of Science:

• Computational Biology
• Bioinformatics
• Evolutionary Biology

Background:

• Multiobjective metaheuristics are increasingly used in computational biology for complex optimization problems.
• Phylogenetic inference, reconstructing evolutionary relationships, is a key area benefiting from these techniques.
• Discrepancies in phylogenetic trees arise from using different optimality principles (e.g., maximum parsimony, maximum likelihood).

Purpose of the Study:

• To propose a novel multiobjective swarm intelligence approach for phylogenetic inference.
• To integrate maximum parsimony and maximum likelihood criteria for a complementary view of phylogenetics.
• To generate a set of phylogenetic trees representing a compromise between different evolutionary principles.

Main Methods:

• Development of a multiobjective swarm intelligence algorithm.
• Utilizing a novel Artificial Bee Colony (ABC) algorithm variant.
• Application to nucleotide datasets for phylogenetic tree reconstruction.

Main Results:

• The proposed method successfully generates a set of compromise phylogenetic trees.
• Experimental results demonstrate the relevance of the approach on various nucleotide datasets.
• Statistical studies validate the proposal against other multiobjective algorithms and current biological methods.

Conclusions:

• The multiobjective Artificial Bee Colony approach offers a valuable complementary perspective in phylogenetic inference.
• This method effectively balances conflicting optimality criteria in reconstructing evolutionary history.
• The findings highlight the potential of advanced metaheuristics in advancing computational biology and evolutionary studies.