Python package for performing multi-state iterative Boltzmann inversion.
copied from cf-post-staging / msibiAn intuitive python API for developing coarse-grain potentials using the multi-state iterative Boltzmann inversion (MSIBI) method. msibi is able to optimize bond-stretching, bond-bending, non-bonded pairs and bonded dihedral potentials where multiple optimizations can be ran in sequence to create a complete coarse grain forcefield. msibi utilizes the HOOMD-Blue simulation engine, and can therefore run on compatible GPU-hardware. The table potentials created by msibi are useable with LAMMPS, GROMACS, HOOMD-Blue, or any simulation engine that supports tabulated potentials.