Finding transition pathways using the string method with swarms of trajectories.

Area of Science:

• Computational chemistry
• Molecular dynamics
• Biophysics

Background:

• Characterizing transition pathways is crucial for understanding complex system dynamics.
• Existing methods like the string method provide frameworks but can be computationally intensive.
• Identifying the most probable transition path (MPTP) is key to elucidating reaction mechanisms.

Purpose of the Study:

• To present a novel, conceptually simple, and easily implementable method for finding transition pathways in complex systems.
• To refine initial trial paths into the MPTP using average dynamic drift.
• To demonstrate the method's efficiency and accuracy on benchmark and complex biological systems.

Main Methods:

• Refining a putative transition path using the average dynamic drift of collective variables.
• Estimating drift on-the-fly with short, unbiased trajectories.
• Iteratively updating the path and distributing computations across multiple nodes.

Main Results:

• Successfully determined the C7eq to C7ax transition pathway for alanine dipeptide, with committor distribution matching previous studies.
• Applied the method to the allosteric conformational change in nitrogen regulatory protein C (NtrC) using over 550 collective variables.
• Observed rapid convergence and a committor distribution peaked at 1/2 near the free energy maximum for NtrC, indicating successful localization of the transition state.

Conclusions:

• The presented method offers an efficient and robust approach for identifying transition pathways in complex multidimensional systems.
• The technique accurately localizes transition states, as validated by its application to both small molecules and large proteins.
• This method has significant potential for advancing the study of molecular mechanisms and conformational changes.