Checklist

[uppittu11]

Complete workflow: mbuild -> foyer -> parmed -> GMSO -> GROMACS

mbuild:

• [x] Small molecules in a box (user can select a molecule using SMILES string or from a pdb files (provided))

foyer:

• [x] Typed with the GAFF package
• [x] Charge assignment with antefoyer
• [x] Return a parmed structure

GMSO:

• [x] Convert the parmed structure to a GMSO Topology
• [x] Show case the way parameters are stored (unyt, expressions)
• [x] Save out to .gro and .top (Need to add whatever form used on GAFF to the accepted_potentials: https://github.com/mosdef-hub/gmso/blob/3d363fc8bde5984b97e7c71ef36e9ea0c49e1e3e/gmso/formats/top.py#L235)

GROMACS:

• [x] Run the simulation of the box for a bit and show energy?

General

• [ ] Add image in the README
• [ ] Switch out molecules and reduce number of molecules
• [x] Make the notebook a binder

[rmatsum836]

A couple other things to resolve:

• Right now the long-range cutoffs are set to 1.2 nm. If someone initializes a box size less than 2.4 nm, GROMACS will fail.

Compounds that don't type:

• 11-dichloroethane
• 13-dichloropropane

[uppittu11]

Yeah. We should replace the cutoff lengths to what GAFF uses. And we should probably do small gas phase systems by 1) restricting the density and n_compounds or keeping the box size constant. If we make sure the density is less than 5 kg/m3 then the box will probably be large enough.

Also, the list of compounds is too long. We should probably keep 10-15 really cool compounds that people would've heard of (like caffeine, cocaine, cholesterol, propane, carbon dioxide, ethanol, aspirin, sucrose, citric acid, ...). It would also make validating all of them easier