Wiki

Clone wiki

MIST / MIST Integrators

This section details the integrators available in the MIST library.

Page Table of Contents

MIST Integrators

MIST is designed to facilitate the addition of new integrators into the packages that it supports. The following is a list of current integrators available and the options available in the input file.

For each integrator we provide the keyword to be used in mist.params to select the choice of integrator as well as any parameters.

Verlet

Velocity Verlet integrator

Keyword: verlet

Parameters: none

Leapfrog

Verlet Leapfrog integrator - NB, velocities are half a step ahead of positions at the end of each step

Keyword: leapfrog

Parameters: none

Langevin

Langevin dynamics using the BAOAB splitting method

Keyword: langevin

Parameters:

  • double langfriction - friction in (ps)^-1 (default = 1.0 ps^-1)
  • double temperature - temperature (K) (default = 300 K)
  • integer seed - for random number generator seed (default = 100)

Yoshida4

4th order symplectic integrator

Keyword: yoshida4

Parameters: none

Yoshida8

8th order symplectic integrator

Keyword: yoshida8

Parameters: none

RK4

4th order Runge-Kutta integrator

Keyword: rk4

Parameters: none

Continuous Tempering

Implementation of continuous tempering developed by Gianpaolo Gobbo.

Keyword: tempering

Parameters:

  • double langfriction - friction in (ps)^-1 (default = 1.0 ps^-1)
  • double langtemp - temperature (K) (default = 300 K)
  • double temp_fact - temperature rescaling factor (default = 0.5)
  • double mass_xi - mass of the additional DOF (default = 0.004)
  • double langfriction_xi - friction of the additional DOF in (ps)^-1 (default = 1.0)
  • double delta - interval of additional DOF for which the coupling function is zero in A (default = 50)
  • double delta2 - interval of additional DOF for which the coupling function reaches the maximum in A (default = 100)
  • double delta-wall - tuning parameter for shape of wall confining additional DOF (default = 50)
  • double hills-height - hill height for metadynamics on addition DOF in kj/mol (default = 7.5)
  • double hills-width - hill width for metadynamics on addition DOF (default = 3.05)
  • integer metadyn-pace - period of hills deposition in time steps (default = 1000)
  • integer save-config-freq - period of saving additional DOF configuration parameter in time steps (default = 500)
  • integer read-bias-file - if set, the bias potential is read from a file meta.0, otherwise it is constructed from scratch
  • double peoffset - potential energy offset (default = 0.0)
  • integer seed - for random number generator seed (default = 100)

TAMD

Implementation of Temperature Accelerated Molecular Dynamics developed by Ralf Banisch. Additional degrees of freedom are coupled to two backbone dihedral angles (phi and psi).

Keyword: tamd

Parameters:

  • double langfriction - friction in (ps)^-1 (default = 1.0 ps^-1)
  • double langtemp - temperature (K) (default = 300 K)
  • double tamd_friction - TAMD friction in (ps)^-1 (default = 10.0 ps^-1)
  • double tamd_temp - TAMD temperature (K) (default = 1500 K)
  • double tamd_mass - TAMD mass ib u*A^2 (default = 168.0 )
  • double tamd_kappa - TAMD kappa parameter in KJ/mol (default = 1163.0)
  • integer tamd_save_freq - period with which to save output in time steps (default = 10)
  • integer seed - for random number generator seed (default = 100)
  • integer phi1, phi2, phi3, phi4 - atom indices defining the first dihedral angle
  • integer psi1, psi2, psi3, psi4 - atom indices defining the first dihedral angle

Continuous Tempering + TAMD

Combined Temperature Accelerated Molecular Dynamics with Continuous Tempering developed by Ralf Banisch. Additional degrees of freedom are coupled to two backbone dihedral angles (phi and psi). The additional DOFs are then heated using the Continuous Tempering Algorithm.

Keyword: tempering_tamd

Parameters:

  • double langfriction - friction in (ps)^-1 (default = 1.0 ps^-1)
  • double langtemp - temperature (K) (default = 300 K)
  • double tamd_friction - TAMD friction in (ps)^-1 (default = 10.0 ps^-1)
  • double tamd_temp - TAMD temperature (K) (default = 1500 K)
  • double tamd_mass - TAMD mass ib u*A^2 (default = 168.0 )
  • double tamd_kappa - TAMD kappa parameter in KJ/mol (default = 1163.0)
  • integer tamd_save_freq - period with which to save output in time steps (default = 10)
  • integer seed - for random number generator seed (default = 100)
  • integer phi1, phi2, phi3, phi4 - atom indices defining the first dihedral angle
  • integer psi1, psi2, psi3, psi4 - atom indices defining the first dihedral angle

The parameters associated with the Continuous Tempering algorithm are currently hard-coded in ContinuousTempering_TAMD.cpp

Simulated Tempering

Implementation of Simulated Tempering with on-the-fly weight determination with temperature control using Langevin Dynamics.

Keyword: simulated_tempering

Parameters:

  • double langfriction - friction in (ps)^-1 (default = 1.0 ps^-1)
  • double temperature - starting temperature (K) (default = 300 K)
  • integer period - number of steps between attempted temperature state switches
  • double[] temperatures - Array of temperature states (K)

Nose-Hoover NPT

Implementation of a Nose-Hoover thermostat and isotropic Nose-Hoover / Parinello-Rahman barostat

Keyword: nose-hoover

  • double temperature - target temperature (K) (default = 300K)
  • integer temp_chain_length - length of the thermostat Nose-Hoover chain (default = 3)
  • integer temp_nsteps_relax - number of timesteps over which the thermostat should relax (default = 100)
  • double pressure - target pressure (Atm) (default = 1 Atm)
  • integer press_chain_length - length of the barostat Nose-Hoover chain (default = 3)
  • integer press_nsteps_relax - number of timesteps over which the barostat should relax (default = 100) temperature", "K");

Sample parameters input file

To run one of the support MD codes with the MIST library you need to provide an input file mist.params containing your integrator choice and any relevant parameters for that integrator. An example parameters file for the Langevin integrator would be:

## Select which integration algorithm to use
integrator langevin

# Parameters for the Langevin integrators
temperature 300
langfriction 1.0

Up: Home; Previous: Running NAMD-Lite with MIST; Next: Constraints with MIST

Updated