Inferring functional units in ion channel pores via relative entropy

Area of Science:

• Computational biology
• Biophysics
• Protein modeling

Background:

• Coarse-grained protein models approximate physical potentials for large biomolecules.
• The relative entropy framework offers a physically sound method for developing and refining these models.
• Current methods focus on parameter fitting, but optimal model topology identification remains a challenge.

Purpose of the Study:

• To extend the relative entropy minimization framework to optimize both model parameters and the mapping to a reduced-dimension topology.
• To develop a more accurate and efficient coarse-grained modeling approach for protein systems.
• To apply and validate the extended framework on anisotropic network models of ion channels.

Main Methods:

• Implementing a two-level optimization procedure based on relative entropy minimization.
• Defining and optimizing the mapping between target protein structures/dynamics and coarse-grained models.
• Utilizing anisotropic network models (ANMs) for ion channel representation.
• Comparing model predictions with experimental data for validation.

Main Results:

• The extended relative entropy minimization successfully optimizes model parameters and identifies optimal reduced-dimension topologies.
• The developed approach provides physically meaningful and accurate representations of protein dynamics.
• Anisotropic network models of ion channels, when optimized with this method, show good agreement with experimental observations.

Conclusions:

• The proposed two-level optimization significantly enhances the capability of coarse-grained protein modeling.
• This framework offers a powerful tool for understanding protein structure-dynamics relationships and function.
• The method holds promise for broader applications in molecular dynamics and drug discovery.