A Novel Approach to Multiple Sequence Alignment Using Multiobjective Evolutionary Algorithm Based on Decomposition.

Area of Science:

• Bioinformatics
• Computational Biology
• Evolutionary Algorithms

Background:

• Multiple sequence alignment (MSA) is crucial for bioinformatics tasks like phylogenetic analysis and motif identification.
• Existing MSA methods lack biological perfection, necessitating novel approaches.

Purpose of the Study:

• To propose and evaluate MOMSA, a new multiple sequence alignment method.
• To treat MSA as a multiobjective optimization problem solved by an evolutionary algorithm.

Main Methods:

• Developed a novel population initialization and mutation operator for the MOMSA algorithm.
• Applied a multiobjective evolutionary algorithm framework based on decomposition.
• Compared MOMSA against evolutionary and progressive alignment methods on BAliBASE 2.0 and 3.0 datasets.

Main Results:

• MOMSA achieved significantly better alignments than VDGA and GAPAM across most test cases.
• MOMSA outperformed IMSA in terms of TC scores.
• MOMSA's alignment quality is comparable to state-of-the-art progressive alignment approaches.

Conclusions:

• MOMSA offers a promising new strategy for multiple sequence alignment.
• The multiobjective optimization approach effectively improves alignment accuracy.
• MOMSA provides competitive performance against established bioinformatics tools.