Improving double-ended transition state searches for soft-matter systems.

Area of Science:

• Computational Chemistry
• Molecular Dynamics
• Biophysics

Background:

• Understanding biomolecular transitions is key for disease mechanisms and molecular engineering.
• Geometry optimization using double-ended transition state searches is a common method.
• Interpolation schemes create bands of transition state candidates between known states.

Purpose of the Study:

• To analyze the accuracy of various interpolation schemes in characterizing biomolecular transition pathways.
• To investigate the impact of different interpolation parameters on pathway representation.
• To determine the necessity of further refinement beyond initial band optimization.

Main Methods:

• Compared multiple interpolation schemes for transition state searches.
• Systematically varied the number of discrete images and spring constants.
• Tested schemes for adjusting spring constants and image distribution (2760 attempts).
• Utilized the doubly-nudged elastic band method for spring constant adjustment.

Main Results:

• Optimized bands do not always accurately depict true transition pathways, especially when low-energy minima are involved.
• Adjustment of spring constants with the doubly-nudged elastic band method improved band quality.
• Image redistribution offered a smaller improvement compared to spring constant adjustment.
• The findings are applicable to diverse molecular and condensed matter systems.

Conclusions:

• Optimized bands are insufficient on their own for describing transition pathways.
• Transition states must be converged, and their connectivity established for reliable pathway characterization.
• Refined methods, particularly adjusting spring constants, enhance the description of molecular transitions.