On closing the inopportune gap with consistency transformation and iterative refinement

Area of Science:

• Bioinformatics
• Computational Biology
• Genomics

Background:

• Multiple Sequence Alignment (MSA) is crucial in bioinformatics.
• Progressive alignment, a common MSA method, faces challenges with gap insertion.
• Consistency Transformation and Iterative Refinement are used to improve MSA quality.

Purpose of the Study:

• To evaluate the alignment quality benefits of Consistency Transformation and Iterative Refinement.
• To compare four classes of MSA heuristics: Progressive, Iterative, Consistency-based, and Consistency-based Progressive with Iterative Refinement.
• To analyze performance across different sequence similarity levels using HOMSTRAD and BAliBASE benchmarks.

Main Methods:

• A tool-independent, technique-oriented analysis of MSA strategies.
• Comparison of 80 MSA aligners across four heuristic classes.
• Statistical evaluation of alignment quality based on benchmark datasets.

Main Results:

• Consistency-based methods statistically outperform others for low-similarity sequences.
• For high-similarity sequences, differences between heuristic classes are less pronounced.
• Iterative Refinement shows limited statistical advantage for progressive aligners.

Conclusions:

• The effectiveness of MSA techniques varies with sequence similarity and problem instances.
• No single heuristic class universally outperforms others.
• Development of integrated MSA frameworks considering multiple techniques is motivated.