Finding multiple reaction pathways via global optimization of action.

Area of Science:

• Computational chemistry
• Computational biology
• Biophysics

Background:

• Global searching for reaction pathways is a significant challenge in computational chemistry and biology.
• Existing methods often rely on local searches due to high computational complexity.
• Identifying multiple diverse pathways is crucial for understanding complex reactions.

Purpose of the Study:

• To present a novel computational approach, Action-CSA, for global optimization of reaction pathways.
• To enable the discovery of multiple diverse reaction pathways connecting specified initial and final states.
• To overcome limitations of local search methods in computational chemistry and biology.

Main Methods:

• Action-CSA utilizes global optimization of the Onsager-Machlup action.
• The conformational space annealing (CSA) method is employed within Action-CSA.
• The approach allows for pathway crossovers and mutations to overcome energy barriers.

Main Results:

• Action-CSA successfully identifies multiple diverse reaction pathways without requiring initial pathway guesses.
• The method efficiently overcomes significant energy barriers.
• Pathway rank order and transition time distributions align well with Langevin dynamics simulations.
• The lowest action folding pathway for FSD-1 matches experimental findings.

Conclusions:

• Action-CSA is an efficient and robust computational tool for studying complex reaction pathways.
• The method is effective for analyzing large-scale conformational changes.
• Action-CSA advances the field of computational chemistry and biology by enabling comprehensive pathway exploration.