Metadynamics

Official Resources

• Homepage: Method - Implemented in PLUMED, CP2K, VASP, etc.
• Documentation: https://www.plumed.org/doc (PLUMED implementation)
• Source Repository: https://github.com/plumed/plumed2 (Reference implementation)
• License: Varies by implementation (LGPL for PLUMED)

Overview

Metadynamics is an enhanced sampling method used to reconstruct free energy surfaces and accelerate rare events in molecular dynamics simulations. It works by adding a history-dependent bias potential (usually Gaussian hills) to selected collective variables (CVs), encouraging the system to explore new regions of phase space.

Scientific domain: Enhanced sampling, free energy calculations, rare events Target user community: Computational chemists, materials scientists, physicists

Theoretical Methods

• Standard Metadynamics
• Well-Tempered Metadynamics (WT-MetaD)
• Bias-Exchange Metadynamics
• Parallel Tempering Metadynamics
• Transition Path Sampling
• Reweighting techniques

Capabilities (CRITICAL)

• Reconstructing free energy landscapes (FES)
• Accelerating transitions between metastable states
• Sampling high-dimensional reaction coordinates
• Estimating equilibrium properties from non-equilibrium runs
• Implemented in: PLUMED (plugin for many codes), CP2K, DESMO-J, and native implementations in some MD codes

Sources: Laio and Parrinello, PNAS 99, 12562 (2002)

Inputs & Outputs

• Input: Definition of collective variables (CVs), Gaussian height/width, deposition rate
• Output: Time series of CVs, HILLS file (list of added biases), Reconstructed Free Energy Surface

Interfaces & Ecosystem

• PLUMED: The de facto standard library for metadynamics
• VASP: Native or PLUMED interface
• CP2K: Strong internal implementation
• GROMACS/LAMMPS/NAMD: Via PLUMED

Workflow and Usage

1. Identify CVs: Choose variables that describe the slow degrees of freedom.
2. Setup: Configure metadynamics parameters (bias factor, sigma, height).
3. Run: Execute MD simulation with metadynamics bias active.
4. Analysis: Sum the deposited Gaussians to obtain the negative free energy surface.
5. Convergence: Check for diffusive behavior in CV space.

Performance Characteristics

• Overhead: Requires calculation of CVs and bias update (usually low overhead)
• Convergence: Depends critically on the choice of CVs
• Scalability: Parallelizable (multiple walkers)

Application Areas

• Crystal nucleation
• Chemical reactions
• Protein folding/unfolding
• Ligand binding
• Phase transitions

Community and Support

• Large community centered around PLUMED
• Active development of new variants (e.g., OPES)
• Theoretical chemistry community

Verification & Sources

Primary sources:

1. PLUMED: https://www.plumed.org/
2. Publication: A. Laio and M. Parrinello, PNAS 99, 12562 (2002)

Confidence: VERIFIED

Verification status: ✅ VERIFIED

• Method: STANDARD
• Documentation: COMPREHENSIVE (via PLUMED)
• Applications: Free energy, enhanced sampling