OpenMM Users Manual and Theory Guide

• 1. Introduction
  • 1.1. Online Resources
  • 1.2. Referencing OpenMM
  • 1.3. Acknowledgments
• 2. The OpenMM Application Layer: Getting Started
  • 2.1. Introduction
  • 2.2. Installing OpenMM
• 3. Running Simulations
  • 3.1. A First Example
  • 3.2. Using AMBER Files
  • 3.3. Using Gromacs Files
  • 3.4. Using CHARMM Files
  • 3.5. The OpenMM-Setup Application
  • 3.6. Simulation Parameters
    • 3.6.1. Platforms
    • 3.6.2. Force Fields
      • 3.6.2.1. Amber14
      • 3.6.2.2. CHARMM36
      • 3.6.2.3. AMOEBA
      • 3.6.2.4. CHARMM Polarizable Force Field
      • 3.6.2.5. Older Force Fields
      • 3.6.2.6. Water Models
    • 3.6.3. Small molecule parameters
      • 3.6.3.1. Small molecule residue template generators
      • 3.6.3.2. Managing force fields with SystemGenerator
    • 3.6.4. AMBER Implicit Solvent
    • 3.6.5. Nonbonded Interactions
      • 3.6.5.1. Nonbonded Forces for AMOEBA
    • 3.6.6. Constraints
    • 3.6.7. Heavy Hydrogens
    • 3.6.8. Integrators
      • 3.6.8.1. Langevin Middle Integrator
      • 3.6.8.2. Langevin Integrator
      • 3.6.8.3. Nosé-Hoover Integrator
      • 3.6.8.4. Leapfrog Verlet Integrator
      • 3.6.8.5. Brownian Integrator
      • 3.6.8.6. Variable Time Step Langevin Integrator
      • 3.6.8.7. Variable Time Step Leapfrog Verlet Integrator
      • 3.6.8.8. Multiple Time Step Integrator
      • 3.6.8.9. Multiple Time Step Langevin Integrator
      • 3.6.8.10. Compound Integrator
    • 3.6.9. Temperature Coupling
    • 3.6.10. Pressure Coupling
    • 3.6.11. Energy Minimization
    • 3.6.12. Removing Center of Mass Motion
    • 3.6.13. Writing Trajectories
    • 3.6.14. Recording Other Data
    • 3.6.15. Saving Simulation Progress and Results
    • 3.6.16. Enhanced Sampling Methods
      • 3.6.16.1. Simulated Tempering
      • 3.6.16.2. Metadynamics
      • 3.6.16.3. Accelerated Molecular Dynamics (aMD)
• 4. Model Building and Editing
  • 4.1. Adding Hydrogens
  • 4.2. Adding Solvent
  • 4.3. Adding a Membrane
  • 4.4. Adding or Removing Extra Particles
  • 4.5. Removing Water
  • 4.6. Saving The Results
• 5. Advanced Simulation Examples
  • 5.1. Simulated Annealing
  • 5.2. Applying an External Force to Particles: a Spherical Container
  • 5.3. Extracting and Reporting Forces (and other data)
  • 5.4. Computing Energies
• 6. Creating Force Fields
  • 6.1. Basic Concepts
    • 6.1.1. Atom Types and Atom Classes
    • 6.1.2. Residue Templates
    • 6.1.3. Forces
  • 6.2. Writing the XML File
    • 6.2.1. <AtomTypes>
    • 6.2.2. <Residues>
    • 6.2.3. <Patches>
    • 6.2.4. Missing residue templates
    • 6.2.5. <HarmonicBondForce>
    • 6.2.6. <HarmonicAngleForce>
    • 6.2.7. <PeriodicTorsionForce>
    • 6.2.8. <RBTorsionForce>
    • 6.2.9. <CMAPTorsionForce>
    • 6.2.10. <NonbondedForce>
    • 6.2.11. <GBSAOBCForce>
    • 6.2.12. <CustomBondForce>
    • 6.2.13. <CustomAngleForce>
    • 6.2.14. <CustomTorsionForce>
    • 6.2.15. <CustomNonbondedForce>
    • 6.2.16. <CustomGBForce>
    • 6.2.17. <CustomHbondForce>
    • 6.2.18. <CustomManyParticleForce>
    • 6.2.19. Writing Custom Expressions
    • 6.2.20. Tabulated Functions
    • 6.2.21. Residue Template Parameters
    • 6.2.22. Including Other Files
  • 6.3. Using Multiple Files
  • 6.4. Extending ForceField
    • 6.4.1. Adding new force types
    • 6.4.2. Adding residue template generators
• 7. The OpenMM Library: Introduction
  • 7.1. What Is the OpenMM Library?
    • 7.1.1. How to get started
    • 7.1.2. License
  • 7.2. Design Principles
  • 7.3. Choice of Language
  • 7.4. Architectural Overview
  • 7.5. The OpenMM Public API
  • 7.6. The OpenMM Low Level API
  • 7.7. Platforms
• 8. Compiling OpenMM from Source Code
  • 8.1. Prerequisites
    • 8.1.1. Get a C++ compiler
      • 8.1.1.1. Mac and Linux: clang or gcc
      • 8.1.1.2. Windows: Visual Studio
    • 8.1.2. Install CMake
    • 8.1.3. Get the OpenMM source code
    • 8.1.4. Other Required Software
  • 8.2. Step 1: Configure with CMake
    • 8.2.1. Build and source directories
    • 8.2.2. Starting CMake
      • 8.2.2.1. Mac and Linux
      • 8.2.2.2. Windows
      • 8.2.2.3. All platforms
  • 8.3. Step 2: Generate Build Files with CMake
    • 8.3.1. Windows
    • 8.3.2. Mac and Linux
  • 8.4. Step 3: Build OpenMM
    • 8.4.1. Windows
    • 8.4.2. Mac and Linux
  • 8.5. Step 4: Install OpenMM
    • 8.5.1. Windows
    • 8.5.2. Mac and Linux
  • 8.6. Step 5: Install the Python API
    • 8.6.1. Windows
    • 8.6.2. Mac and Linux
  • 8.7. Step 6: Test your build
    • 8.7.1. Windows
    • 8.7.2. Mac and Linux
  • 8.8. Building the Documentation (Optional)
    • 8.8.1. User Guide and Developer Guide
    • 8.8.2. Python and C++ API Documentation
• 9. OpenMM Tutorials
  • 9.1. Example Files Overview
  • 9.2. Running Example Files
    • 9.2.1. Visual Studio
      • 9.2.1.1. Determining the platform being used
      • 9.2.1.2. Visualizing the results
    • 9.2.2. Mac OS X/Linux
      • 9.2.2.1. Determining the platform being used
      • 9.2.2.2. Visualizing the results
      • 9.2.2.3. Compiling Fortran and C examples
  • 9.3. HelloArgon Program
    • 9.3.1. Including OpenMM-defined functions
    • 9.3.2. Running a program on GPU platforms
    • 9.3.3. Running a simulation using the OpenMM public API
    • 9.3.4. Error handling for OpenMM
    • 9.3.5. Writing out PDB files
    • 9.3.6. HelloArgon output
  • 9.4. HelloSodiumChloride Program
    • 9.4.1. Simple molecular dynamics system
    • 9.4.2. Interface routines
      • 9.4.2.1. Units
      • 9.4.2.2. Lennard-Jones potential
      • 9.4.2.3. Opaque handle MyOpenMMData
      • 9.4.2.4. myInitializeOpenMM
      • 9.4.2.5. myGetOpenMMState
      • 9.4.2.6. myStepWithOpenMM
      • 9.4.2.7. myTerminateOpenMM
  • 9.5. HelloEthane Program
• 10. Platform-Specific Properties
  • 10.1. OpenCL Platform
  • 10.2. CUDA Platform
  • 10.3. CPU Platform
  • 10.4. Determinism
• 11. Using OpenMM with Software Written in Languages Other than C++
  • 11.1. C API
    • 11.1.1. Mechanics of using the C API
    • 11.1.2. Mapping from the C++ API to the C API
    • 11.1.3. Exceptions
    • 11.1.4. OpenMM_Vec3 helper type
    • 11.1.5. Array helper types
      • 11.1.5.1. OpenMM_DoubleArray
      • 11.1.5.2. OpenMM_StringArray
      • 11.1.5.3. OpenMM_Vec3Array
      • 11.1.5.4. OpenMM_BondArray
      • 11.1.5.5. OpenMM_ParameterArray
  • 11.2. Fortran 95 API
    • 11.2.1. Mechanics of using the Fortran API
    • 11.2.2. Mapping from the C++ API to the Fortran API
    • 11.2.3. OpenMM_Vec3 helper type
    • 11.2.4. Array helper types
      • 11.2.4.1. OpenMM_DoubleArray
      • 11.2.4.2. OpenMM_StringArray
      • 11.2.4.3. OpenMM_Vec3Array
      • 11.2.4.4. OpenMM_BondArray
      • 11.2.4.5. OpenMM_ParameterArray
  • 11.3. Python API
    • 11.3.1. Mapping from the C++ API to the Python API
    • 11.3.2. Mechanics of using the Python API
    • 11.3.3. Units and dimensional analysis
      • 11.3.3.1. Why does the Python API include units?
      • 11.3.3.2. Quantities, units, and dimensions
      • 11.3.3.3. Units examples
      • 11.3.3.4. Arithmetic with units
      • 11.3.3.5. Atomic scale mass and energy units are “per amount”
      • 11.3.3.6. SI prefixes
      • 11.3.3.7. Converting to different units
      • 11.3.3.8. Lists, tuples, vectors, numpy arrays, and Units
• 12. Examples of OpenMM Integration
  • 12.1. GROMACS
  • 12.2. TINKER-OpenMM
• 13. Testing and Validation of OpenMM
  • 13.1. Description of Tests
    • 13.1.1. Unit tests
    • 13.1.2. System tests
    • 13.1.3. Direct comparisons between OpenMM and other programs
  • 13.2. Test Results
    • 13.2.1. Comparison to Reference Platform
    • 13.2.2. Energy Conservation
    • 13.2.3. Comparison to Gromacs
• 14. AMOEBA Plugin
  • 14.1. OpenMM AMOEBA Supported Forces and Options
    • 14.1.1. Supported Forces and Options
    • 14.1.2. Supported Integrators
  • 14.2. TINKER-OpenMM
    • 14.2.1. Building TINKER-OpenMM (Linux)
    • 14.2.2. Using TINKER-OpenMM
      • 14.2.2.1. Available outputs
      • 14.2.2.2. Setting the frequency of output data updates
      • 14.2.2.3. Specify the GPU board to use
      • 14.2.2.4. Running comparison tests between TINKER and OpenMM routines
      • 14.2.2.5. Turning off OpenMM / Reverting to TINKER routines
  • 14.3. OpenMM AMOEBA Validation
• 15. Ring Polymer Molecular Dynamics (RPMD) Plugin
• 16. Drude Plugin
• 17. The Theory Behind OpenMM: Introduction
  • 17.1. Overview
  • 17.2. Units
  • 17.3. Physical Constants
• 18. Standard Forces
  • 18.1. HarmonicBondForce
  • 18.2. HarmonicAngleForce
  • 18.3. PeriodicTorsionForce
  • 18.4. RBTorsionForce
  • 18.5. CMAPTorsionForce
  • 18.6. NonbondedForce
    • 18.6.1. Lennard-Jones Interaction
    • 18.6.2. Coulomb Interaction Without Cutoff
    • 18.6.3. Coulomb Interaction With Cutoff
    • 18.6.4. Coulomb Interaction With Ewald Summation
    • 18.6.5. Coulomb Interaction With Particle Mesh Ewald
    • 18.6.6. Lennard-Jones Interaction With Particle Mesh Ewald
  • 18.7. GBSAOBCForce
    • 18.7.1. Generalized Born Term
    • 18.7.2. Surface Area Term
  • 18.8. GayBerneForce
  • 18.9. AndersenThermostat
  • 18.10. MonteCarloBarostat
  • 18.11. MonteCarloAnisotropicBarostat
  • 18.12. MonteCarloMembraneBarostat
  • 18.13. CMMotionRemover
  • 18.14. RMSDForce
• 19. Custom Forces
  • 19.1. CustomBondForce
  • 19.2. CustomAngleForce
  • 19.3. CustomTorsionForce
  • 19.4. CustomNonbondedForce
  • 19.5. CustomExternalForce
  • 19.6. CustomCompoundBondForce
  • 19.7. CustomCentroidBondForce
  • 19.8. CustomManyParticleForce
  • 19.9. CustomGBForce
  • 19.10. CustomHbondForce
  • 19.11. CustomCVForce
  • 19.12. Writing Custom Expressions
  • 19.13. Setting Parameters
  • 19.14. Parameter Derivatives
• 20. Integrators
  • 20.1. VerletIntegrator
  • 20.2. LangevinIntegator
  • 20.3. LangevinMiddleIntegrator
  • 20.4. NoseHooverIntegrator
  • 20.5. BrownianIntegrator
  • 20.6. VariableVerletIntegrator
  • 20.7. VariableLangevinIntegrator
  • 20.8. CustomIntegrator
• 21. Other Features
  • 21.1. Periodic Boundary Conditions
  • 21.2. LocalEnergyMinimizer
  • 21.3. XMLSerializer
  • 21.4. Force Groups
  • 21.5. Virtual Sites
  • 21.6. Random Numbers with Stochastic Integrators and Forces
• 22. Bibliography

Portions copyright (c) 2008-2017 Stanford University and the Authors

Contributors: Kyle Beauchamp, Christopher Bruns, John Chodera, Peter Eastman, Mark Friedrichs, Joy P. Ku, Tom Markland, Vijay Pande, Randy Radmer, Michael Sherman, Jason Swails, Lee-Ping Wang

Permission is hereby granted, free of charge, to any person obtaining a copy of this document (the “Document”), to deal in the Document without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Document, and to permit persons to whom the Document is furnished to do so, subject to the following conditions:

This copyright and permission notice shall be included in all copies or substantial portions of the Document.

THE DOCUMENT IS PROVIDED “AS IS”, WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS, CONTRIBUTORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE DOCUMENT OR THE USE OR OTHER DEALINGS IN THE DOCUMENT.