ORION: a web server for protein fold recognition and structure prediction using evolutionary hybrid profiles

Area of Science:

• Computational biology
• Structural bioinformatics
• Protein structure prediction

Background:

• Comparative modeling is an efficient method for protein structure prediction.
• Accurate template identification is crucial for successful comparative modeling.

Purpose of the Study:

• To introduce ORION, a web server for protein structure prediction using comparative modeling.
• To present the novel profile-profile approach employed by ORION for template identification.
• To highlight the recent improvements in ORION's performance.

Main Methods:

• ORION utilizes a profile-profile approach combining sequence and structure evolution information.
• Structural features like solvent accessibility and Protein Blocks are encoded into profiles.
• Homologous protein structures are searched in databases like PDB, SCOP, and HOMSTRAD.
• Protein 3D structures are modeled using MODELLER and quality estimation measures.

Main Results:

• ORION identifies suitable templates using an original profile-profile approach.
• The web server integrates sequence and structure evolution information for enhanced predictions.
• Recent improvements have increased ORION's result quality by 5%.
• ORION can predict protein structures from a single sequence within minutes.

Conclusions:

• ORION provides an efficient and improved web server for protein structure prediction via comparative modeling.
• The novel template identification strategy enhances the accuracy of predicted protein structures.
• ORION is a valuable tool for researchers needing rapid and reliable protein structure models.