Gaussian network model can be enhanced by combining solvent accessibility in proteins.

Area of Science:

• Computational biology
• Protein dynamics analysis
• Biophysics

Background:

• The Gaussian network model (GNM) is a fundamental coarse-grained method for studying protein dynamics and function.
• Improving residue flexibility modeling in GNM requires innovative approaches to its Kirchhoff matrix.
• Previous work showed relative solvent accessibility (RSA) enhances linear regression models for residue flexibility.

Purpose of the Study:

• To develop an improved parameter-free GNM by incorporating local relative solvent accessibility (RSA).
• To enhance the accuracy of protein residue flexibility predictions.
• To introduce a novel GNM variation, termed RpfGNM, for advanced protein dynamics analysis.

Main Methods:

• Modified the Kirchhoff matrix of the parameter-free GNM by including local relative solvent accessibility (RSA) information.
• Utilized particle swarm optimization to estimate undetermined parameters in the new Kirchhoff matrix.
• Validated the RpfGNM using training, independent, and molecular dynamics simulation datasets.

Main Results:

• The proposed RSA-based parameter-free GNM (RpfGNM) demonstrated significantly increased average correlation coefficients.
• RpfGNM outperformed the standard parameter-free GNM in predicting residue flexibility across multiple datasets.
• Empirical results confirm the effectiveness of integrating protein structural properties into GNM variations.

Conclusions:

• Integrating local relative solvent accessibility into GNM offers a powerful strategy for enhancing protein flexibility modeling.
• The RpfGNM represents a significant advancement in accurately predicting protein residue flexibility.
• Variations of classical GNM incorporating additional structural properties are promising for future protein dynamics research.