A Comparison of Three Perturbation Molecular Dynamics Methods for Modeling Conformational Transitions

Area of Science:

• Computational biology
• Biophysics
• Molecular modeling

Background:

• Targeted, steered, and biased molecular dynamics (MD) are crucial for studying biomolecular transitions.
• These methods use external perturbations along a progress variable to guide conformational changes.

Purpose of the Study:

• To compare targeted, steered, and biased MD methods for generating biomolecular transition paths.
• To investigate the influence of simulation parameters on path generation for protein unfolding and kinase conformational changes.

Main Methods:

• Applied targeted, steered, and biased molecular dynamics simulations.
• Calculated transition pathways for protein A's B domain unfolding and Src kinase Lyn's catalytic domain transition.
• Varied simulation parameters: progress variable, simulation length, and biasing force constant.
• Computed potentials of mean force (PMF) to estimate path probabilities.

Main Results:

• Targeted, steered, and biased MD methods produced similar transition paths for a given progress variable.
• The choice of progress variable had a stronger influence on path generation than other simulation parameters.
• Lower PMF values were observed with the lowest biasing force constant in biased molecular dynamics (BMD).

Conclusions:

• The choice of progress variable is critical for accurately modeling biomolecular transitions using perturbation-based MD methods.
• Biased molecular dynamics, particularly with optimized force constants, can effectively generate reliable transition pathways.