A low-complexity add-on score for protein remote homology search with COMER

Area of Science:

• Bioinformatics
• Computational Biology
• Structural Bioinformatics

Background:

• Protein sequence alignment is crucial for comparative modeling and protein structure prediction.
• Profile-profile alignment is a sensitive method for homology detection but requires enhancement.

Purpose of the Study:

• To improve the quality and sensitivity of profile-profile alignments.
• To refine profile-profile substitution scores for better homology detection.

Main Methods:

• Development of a novel add-on score for profile-profile substitution scores.
• Comprehensive evaluation of the new score's impact on alignment quality and sensitivity.
• Implementation within the COMER (COmparative Modeling by Entropy and REgression) method.

Main Results:

• The new add-on score statistically significantly improves alignment quality.
• The score enhances the sensitivity of homology detection.
• The score possesses near-optimal computational complexity for easy integration.

Conclusions:

• The refined substitution score effectively boosts profile-profile alignment performance.
• The COMER method, enhanced with the new score, offers improved homology detection.
• The score's efficiency facilitates its adoption in various profile-profile alignment tools.