Improving prediction accuracy in chimeric proteins with windowed multiple sequence alignment.

Area of Science:

• Computational Biology
• Structural Biology
• Bioinformatics

Background:

• Protein structure prediction is crucial for understanding protein function.
• Co-evolving residue pairs, identified via multiple sequence alignment (MSA), signal spatial proximity and are key to accurate predictions.
• AlphaFold utilizes MSAs for high-accuracy protein structure prediction.

Purpose of the Study:

• To investigate the impact of fused or chimeric protein structures on prediction accuracy.
• To determine if modified MSA strategies can improve the prediction of fused protein structures.

Main Methods:

• Analyzing prediction accuracy of structured peptides appended to scaffold proteins.
• Comparing MSAs of individual components versus a combined MSA for fused proteins.
• Evaluating the effect of windowed MSA on prediction accuracy.

Main Results:

• Structured peptides appended to scaffold proteins significantly reduce prediction accuracy.
• Appending the individual peptide's MSA to the scaffold protein's MSA restores prediction accuracy.
• The windowed MSA approach shows promise for predicting fused protein structures.

Conclusions:

• Terminal peptide additions negatively impact protein structure prediction accuracy.
• Integrating component-specific MSAs is vital for accurate chimeric protein structure prediction.
• Windowed MSA is a valuable technique for predicting the structure of fused proteins.