ORION

ORION performs fold recognition to improve protein 3D structure prediction by comparing hybrid profiles that combine sequence and structure evolutionary information.

Key Features:

• Hybrid Profiles: Integrates evolutionary information from both protein sequence and protein structure to improve remote homology detection.
• Structural Alphabet — Protein Blocks (PBs): Encodes local conformations using a 16-state structural alphabet (Protein Blocks) for one-dimensional description of structure.
• Pairwise Profile Comparison: Compares hybrid profiles in a pairwise manner to identify compatible structural templates.
• Benchmark Performance: Systematically validated against PSI-BLAST and HHsearch on CASP 8, 9, and 10 targets and detects approximately 10% more templates at SCOP fold and superfamily levels.

Scientific Applications:

• Structural Biology: Improves template detection for template-based modeling to support accurate protein 3D structure prediction.
• Drug Discovery: Enhances structure predictions used to inform drug discovery efforts that rely on protein structure.
• Functional Annotation: Enables more reliable functional annotation of proteins through improved structural inference.

Methodology:

Performs pairwise comparison of hybrid profiles that encapsulate evolutionary information from sequence and structure and encodes local structural conformations using a 16-state Protein Blocks (PBs) structural alphabet.