ICOSA: A Distance-Dependent, Orientation-Specific Coarse-Grained Contact Potential for Protein Structure Modeling

Area of Science:

• Computational Biology
• Structural Bioinformatics
• Biophysics

Background:

• Protein folding and stabilization are critically influenced by the spatial arrangement of contacting residue pairs.
• Accurate evaluation of inter-residue interactions is essential for predicting protein structures.

Purpose of the Study:

• To develop a novel knowledge-based, coarse-grained contact potential named ICOSA.
• To correlate inter-residue contact distance and orientation for improved pair-wise interaction evaluation.

Main Methods:

• Development of ICOSA using icosahedral local coordinates to analyze residue contact distributions.
• Extension of finite ideal gas reference state theory to icosahedral coordinates.
• Incorporation of long-range contact interactions for enhanced sensitivity and statistical rigor.

Main Results:

• ICOS A, utilizing only backbone atom information, demonstrates performance comparable to all-atom potentials (Rosetta, DFIRE, I-TASSER, OPUS) in distinguishing native from misfolded protein conformations.
• ICOS A surpasses widely used coarse-grained potentials in accuracy.
• ICOS A achieves performance comparable to fine-grained potentials in identifying protein models from the CASP10 benchmark.

Conclusions:

• ICOS A offers a computationally efficient yet accurate method for protein structure prediction.
• The potential's ability to integrate distance and orientation information provides a robust approach to modeling inter-residue interactions.
• ICOS A represents a significant advancement in coarse-grained modeling for structural bioinformatics.