State of the art: refinement of multiple sequence alignments

Area of Science:

• Computational biology
• Bioinformatics
• Structural biology

Background:

• Accurate protein multiple sequence alignments are vital for computational biology.
• Existing alignment strategies have limitations, leading to inaccuracies.
• High-quality alignments are increasingly important for large-scale, high-throughput analyses.

Purpose of the Study:

• To compare the performance of protein alignment refinement algorithms.
• To evaluate the accuracy and efficiency of different refinement methods.
• To introduce and assess a novel refinement approach, REFINER.

Main Methods:

• Extensive comparison of alignment refinement algorithm performance.
• Evaluation using 3D structure-based alignments from the BAliBASE benchmark database.
• Validation with manually curated, high-quality alignments from the Conserved Domain Database (CDD).

Main Results:

• REFINER, a method using conserved regions as constraints, was compared to other refinement algorithms.
• Performance comparison revealed that REFINER offers modest improvements in alignment quality.
• The refinement method generally produces higher-scoring alignments without degrading well-conserved regions.

Conclusions:

• Protein alignment refinement can enhance existing methods.
• The REFINER method demonstrates improved performance in refining protein sequence alignments.
• REFINER provides a valuable tool for improving the accuracy of large-scale protein alignment datasets.