A computational pipeline for protein structure prediction and analysis at genome scale

Area of Science:

• Computational Biology
• Structural Bioinformatics

Background:

• Experimental methods struggle to match the pace of protein sequence generation.
• Computational protein structure prediction offers scalable solutions for structural characterization.

Purpose of the Study:

• To develop an automated pipeline for large-scale protein structure prediction.
• To integrate diverse computational tools for comprehensive structural analysis.

Main Methods:

• The pipeline utilizes the PROSPECT threading-based system as its core component.
• It incorporates tools for domain identification, signal peptide analysis, protein triage (membrane/globular), fold recognition, and atomic model generation.
• A pipeline manager automates workflow based on protein characteristics.

Main Results:

• An automated, web-accessible pipeline for protein structure prediction was implemented.
• The system integrates multiple tools for comprehensive structural analysis.
• Genome-scale predictions were successfully performed on *Caenorhabditis elegans*, *Pyrococcus furiosus*, and cyanobacterial genomes.

Conclusions:

• The automated pipeline enhances the efficiency of protein structure prediction.
• This computational approach complements experimental techniques for large-scale structural biology.
• The pipeline is available for broader scientific use.