Prediction of perturbed proton transfer networks

Area of Science:

• Computational Biology
• Biophysics
• Biochemistry

Background:

• Proton transfer through translocating channels is crucial for biological processes.
• Accurate modeling requires sampling various protonation states and conformations, which is computationally expensive.
• Existing transition network calculations demand significant resources for each system change.

Purpose of the Study:

• To develop a computationally efficient method for predicting proton transfer pathways.
• To reduce the computational cost associated with transition network calculations for perturbed systems.
• To maintain the accuracy of important network properties during prediction.

Main Methods:

• Developed a prediction method based on extensive coarse-graining of transition networks.
• Utilized the minimum spanning tree (MST) of an unperturbed network as an initial guess.
• Employed sensitivity analysis of the MST to refine predictions for perturbed networks.

Main Results:

• The prediction method achieved up to an 80% reduction in calculation costs in a model system.
• Key network properties were preserved in most predictions of perturbed transition networks.
• The MST efficiently connects network nodes with minimal edge weights, forming a core sub-network.

Conclusions:

• The proposed prediction method offers a significant speed-up for transition network calculations.
• This approach provides a computationally feasible way to study complex proton transfer dynamics.
• The method is valuable for analyzing systems with unsampled protonation states or conformational changes.