Sampling saddle points on a free energy surface.

Area of Science:

• Computational Chemistry
• Biophysics
• Materials Science

Background:

• Saddle points on free energy surfaces represent transition states, vital for understanding molecular transitions.
• Finding these states is challenging in complex systems with many degrees of freedom.

Purpose of the Study:

• To develop an algorithm for identifying saddle points on high-dimensional free energy surfaces in real-time.
• To enable the study of transition states without prior knowledge of the free energy function.

Main Methods:

• Utilizes a heterogeneous multi-scale method combining a macro-scale solver (gentlest ascent dynamics) with micro-scale computations (molecular dynamics).
• Calculates forces and Hessian values on-the-fly for high-dimensional free energy surfaces.

Main Results:

• Successfully identified saddle points for alanine dipeptide isomerization using two variables.
• Mapped transition pathways for alanine decamer beta-hairpin formation using 20 variables, revealing a network of minima and saddle points.

Conclusions:

• The developed algorithm effectively handles high-dimensional free energy surfaces and coarse-grained variables.
• Provides a novel approach to visualize and understand complex molecular dynamics and conformational landscapes.