Test container to display in-progress machinery for using amortized inference to refit the C-C LJ parameters (to scale up) from the smirnoff99Frosst forcefield. We are training solely on molar volume and heat of vaporization from cyclohexane.
/python-scripts: contains all python scripts and ipython notebooks demonstrating making estimates of properties usingpymbar, constructing Gaussian process regressions withgpyand implementing MCMC usingemceein order to sample from a posterior distribution of forcefield parameters./reference: contains the state data and coordinate trajectories (for both liquid and gas phase simulations) for a sample reference state./simulation-data: contains observables calculated by simulation at ~120 different parameter states in order to compare to MBAR./monomers: contains mol2 files for running the MBAR analyses./packmol_boxes: contains packmol boxes in a pdb format for running the MBAR analyses.