mquevill/ssages-public

Software Suite for Advanced General Ensemble Simulations

 

SSAGES (Software Suite for Advanced General Ensemble Simulations) is an open-source, engine agnostic, C++11 based advanced sampling package. It is designed to be easy to use, extendable and extremely versatile. It is currently pre-beta, meaning that there are many rough edges, but we are working rapidly to expand its features and fix any bugs. Keep an eye on this page for future updates and see below on how to contribute!

What's New (v0.9.2)

• Critical bug fix for patching Eigen with GROMACS
• Correction to gradient of AngleCV
• Additional checks and docs for multiwalker loggers
• Support newer LAMMPS versions
• Minor source cleanup

To view the full changelog history, refer to HISTORY.

Features

SSAGES currently works with multiple molecular dynamics engines. It contains a variety of collective variables (CVs) and advanced sampling methods.

Highlights

• Engine agnostic framework
• Simple JSON input file syntax
• Easy to add new CVs
• Easy to add new methods
• Much more!

Engines

• GROMACS 5.1.x, 2016.x, 2018.x
• LAMMPS (Most recent versions)
• OpenMD (2.5+)
• QBox (1.63+)

CVs

• Artificial Neural Network (as a function of group positions)
• Atom group coordinate
• Atom group position
• Atom group separation
• Bend angle
• Box volume
• Components of gyration tensor
• Pairwise kernel (coordination number, nearest neighbors)
• Polymer Rouse modes
• Root-mean-square deviation (RMSD)
• Secondary structure (alpha, anti/parallel beta sheet) RMSD
• Torsional angle

Methods

• Adaptive biasing force
• Artificial neural network sampling
• Basis function sampling
• Combined Force–Frequency Sampling
• Metadynamics
• Umbrella sampling
• Finite temperature string
• Nudged elastic band
• Swarm of trajectories
• Forward flux sampling

Installation

The first step is to clone the repository locally.

$ git clone https://github.com/SSAGESproject/SSAGES.git

SSAGES uses a CMake build system. It also requires the use of a support MD engine. For example, to compile with LAMMPS, execute the following

$ cd SSAGES
$ mkdir build && cd build
$ cmake -DLAMMPS_SRC=/path/to/lammps/src ..
$ make

This will build a SSAGES executable which will reside in the build directory.

If you want to use a specific compiler (or if your default compiler is not supported), set the C and C++ compilers with CMAKE_C_COMPILER and CMAKE_CXX_COMPILER, respectively. For example, to use gcc/g++, replace the CMake command with

$ cmake -DLAMMPS_SRC=/path/to/lammps/src -DCMAKE_C_COMPILER=gcc -DCMAKE_CXX_COMPILER=g++ ..

If you want to compile and run unit and integration tests, replace the cmake command in the example above with

$ cmake -DLAMMPS_SRC=/path/to/lammps/src -DBUILD_TESTS=ON ..

MPI

A requisite underlying MPI library also required to run SSAGES. On recent Debian based systems using OpenMPI, the requirement can be installed via:

$ sudo apt-get install libopenmpi-dev openmpi-bin

For more detail on the build system, please check the documentation.

To build the documentation, refer to Documentation README.

Known issues

SSAGES is currently in pre-beta. That means there may be known issues that are not yet resolved. Major issues are listed here.

• Restarts are not fully functioning for all methods.

Contributing

Feel free to fork this project on GitHub. Any pull-requests, feature requests or other form of contributions are welcome.