An approach to improving multiple alignments of protein sequences using predicted secondary structure

Area of Science:

• Bioinformatics
• Computational Biology
• Structural Biology

Background:

• Multiple sequence alignment (MSA) is crucial for understanding protein evolution and function.
• Accurate MSAs are essential for reliable downstream analyses, including secondary structure prediction.
• Existing MSA methods can be limited by sequence-only information.

Purpose of the Study:

• To develop an improved method for multiple sequence alignment using publicly available tools.
• To leverage predicted protein secondary structure information to enhance MSA accuracy.
• To assess the impact of improved MSAs on subsequent secondary structure prediction.

Main Methods:

• A novel method for MSA was developed, incorporating predicted secondary structure and simplified amino acid descriptions.
• Public-domain software and methods were utilized to ensure reproducibility.
• The accuracy of the generated alignments was evaluated against structural alignments.

Main Results:

• The proposed method improved MSA accuracy by 5–14% compared to full sequence profile alignments.
• Alignments generated by this method led to more accurate secondary structure predictions.
• A 6% improvement in three-state (helix, sheet, coil) secondary structure prediction was observed.

Conclusions:

• Integrating predicted secondary structure significantly enhances multiple sequence alignment accuracy.
• Improved MSAs lead to more reliable protein secondary structure predictions.
• The method is accessible, utilizing public-domain software and providing necessary files for replication.