PC_ali: a tool for improved multiple alignments and evolutionary inference based on a hybrid protein sequence and...

Area of Science:

• Bioinformatics
• Computational Biology
• Evolutionary Biology

Background:

• Multiple sequence alignments (MSA) are critical for evolutionary inference but challenging for distantly related proteins.
• Protein structure is more conserved than sequence, suggesting structure-based alignments could aid distant homolog analysis.
• Existing similarity measures often lack a complete, unbiased view of protein conservation.

Purpose of the Study:

• To develop a novel hybrid score for protein sequence and structure similarity.
• To create a tool (PC_ali) for constructing improved multiple sequence alignments (MSA).
• To enhance the accuracy of evolutionary divergence inference for proteins.

Main Methods:

• Examined four sequence and structure-based protein similarity measures.
• Developed a hybrid similarity score (PC_sim) and divergence measure (PC_div) using principal component analysis.
• Implemented the PC_ali program for de novo or modified MSA construction using PC_sim and PC_div.

Main Results:

• PC_sim and PC_div showed strong correlations with evolutionary divergences.
• PC_ali achieved higher aligned fraction and structural scores compared to state-of-the-art tools.
• PC_ali demonstrated high similarity to reference MSAs and improved sequence identity scores.

Conclusions:

• The PC_sim score and PC_ali tool offer a more accurate approach to protein MSA construction.
• Integrating sequence and structure information via PC_sim enhances evolutionary inference.
• PC_ali represents a significant advancement in bioinformatics tools for evolutionary analysis.