Global Free-Energy Landscapes as a Smoothly Joined Collection of Local Maps

Area of Science:

• Computational chemistry and physics
• Molecular dynamics simulations
• Enhanced sampling techniques

Background:

• Enhanced sampling methods are crucial for simulating slow processes in computational chemistry and physics.
• Current methods face limitations due to the need for a small number of collective variables (CVs).
• Inefficient sampling arises from irrelevant CVs or poorly described slow degrees of freedom.

Purpose of the Study:

• To introduce a new biasing method, Adaptive Topography of Landscape for Accelerated Sampling (ATLAS), designed to overcome limitations of existing enhanced sampling techniques.
• To enable efficient sampling in high-dimensional collective variable spaces.
• To improve the accuracy and efficiency of molecular simulations for complex systems.

Main Methods:

• ATLAS employs a divide-and-conquer strategy, partitioning the high-dimensional CV space into basins.
• Each basin is described by an automatically determined, low-dimensional set of variables.
• A well-tempered metadynamics-like bias is applied locally within basins, controlled by indicator functions.
• The unbiased Boltzmann distribution is recovered via reweighting.

Main Results:

• ATLAS successfully handles a large number of CVs, overcoming a key limitation of traditional methods.
• The method allows for efficient exploration of complex free-energy landscapes.
• Iterative updates of the free-energy landscape decomposition enable discovery of new metastable states.

Conclusions:

• ATLAS represents a significant advancement in enhanced sampling, offering improved efficiency and applicability to complex problems.
• The method facilitates straightforward evaluation of conformational and thermodynamic properties.
• ATLAS provides a robust framework for advancing molecular simulations in chemistry and physics.