GOMC, an abbreviation of GPU Optimized Monte Carlo, is an open-source library for simulating molecular systems using the Metropolis Monte Carlo algorithm. The software has been written in object oriented C++, and uses OpenMP and NVIDIA CUDA to allow for execution on multi-core CPU and GPU architectures.
This GOMC Python interface was developed to automatize file handling, simulation launching and data analysis. This allows for simulation pipelines handled by the computer. Other than that, the wrapper does not extend the functionality of GOMC.
If you already have a GOMC input script, in.conf, this can be read into a GOMC object by
from gomc_wrapper import read
gomc = read("in.conf")Then, parameters can be changed by the set method:
gomc.set("Temperature", 298.15)When one is done editing the configuration file, one usually either wants to write a new GOMC input file or run GOMC with the desired configurations. In the former case, one calls the write method:
gomc.write("in.conf.new", verbose=True)In the latter case, one calls the run method:
gomc.run(num_procs=4, gomc_exec="GOMC_CPU_NVT")A GOMC object might also be created from scratch. Then, all the required parameters need to be set. The standard way of doing this, is to use the set method:
from gomc_wrapper import GOMC
gomc = GOMC()
gomc.set("GEMC", "NVT")
gomc.set("Coordinates", 0, "BUTAN_BOX_0.pdb")However, for some operations it might be more convenient to use one of the specialized set-methods add_box, set_steps or set_prob. add_box lets you add a box to the system:
gomc.add_box(coordinates="BUTAN_BOX_0.pdb",
structure="BUTAN_BOX_0.psf",
hmatrix=[[1, 0, 0], [0, 1, 0], [0, 0, 1]])set_steps lets you set all numbers of steps in one call:
gomc.set_steps(run=1000000, eq=100000, adj=1000)set_prob lets you specify the transition probabilities:
gomc.set_prob(dis=0.6, rot=0.3, swap=0.08, vol=0.02)- Write a simple PSF generator