Coupling with SSAGES

DASH was contructed to efficiently couple to the advanced sampling engine SSAGES. This allows for advanced sampling methods such as Basis Function sampling, Adaptive Bias Force sampling, Green’s function sampling, umbrella sampling, and more to be performed at GPU speeds with minimal overhead.

First, SSAGES must be compiled with DASH.

cd path/to/ssages/directory
mkdir build
cd build
cmake .. -DDASH_SRC=/path/to/dash
make

We also need to compile dash on its own in order to generate the python library

Once both DASH and SSAGES are compiled, ensure that libSim.so is in your LD_LIBRARY_PATH environment variable, and you’re ready to go.

SSAGES looks for two functions within your DASH python script - a setupSimulation function which initializes the simulation state and a runSimulation function which takes as an argument a number of timesteps, and must run the simulation for that long.

Example

#import DASH as normal

def setupSimulation():
    state = State()
    #create fixes, add atoms, prepare simulation state


    #function must return the simulation state
    return state

#SSAGES calls this function using the state you returned in setupSimulation
#SSAGES has internally appended a special fix which applies
#operations required for you advanced sampling technique
def runSimulation(state, numTurns):
    integ = IntegratorVerlet(state)
    nvt = FixNoseHoover(state, handle='myThermo', groupHandle='all', temp=300, timeConstant=100)
    state.activateFix(nvt)
    integ.run(numTurns)
    state.deactivateFix(nvt)