Generation of Pathway Signatures by Combining Normal Modes with Weighted Ensemble Simulations.

Area of Science:

• Computational biology
• Molecular dynamics simulations
• Biophysics

Background:

• Molecular dynamics (MD) simulations offer atomic-level insights but are computationally expensive due to system size and timescale limitations.
• Enhanced sampling methods exist but often introduce energetic biases or lack effective progress coordinates.
• Protein normal modes, derived from 3D structure, represent intrinsic dynamics but integrating them into MD for unbiased pathways is challenging.

Purpose of the Study:

• To develop an unbiased and efficient method for generating molecular pathways using protein dynamics.
• To establish a generalizable progress coordinate for enhanced sampling simulations.
• To overcome limitations of existing enhanced sampling techniques in molecular dynamics.

Main Methods:

• Utilized adaptive anisotropic network model (AANM) to generate conformations along protein normal modes.
• Integrated normal mode-derived conformations as a progress coordinate in weighted ensemble simulations.
• Applied the method to protein systems without requiring prior knowledge of their dynamics.

Main Results:

• Demonstrated that normal mode-derived conformations provide a physical and intuitive progress coordinate.
• Achieved a significant boost in simulation efficiency compared to traditional methods.
• Successfully generated continuous, energetically unbiased pathways for protein systems.

Conclusions:

• The combination of normal modes and adaptive anisotropic network model offers a generalizable and efficient approach for molecular dynamics simulations.
• This method overcomes key challenges in enhanced sampling, providing unbiased pathways.
• The technique holds promise for studying a wide range of protein systems computationally.