shirtsgroup / InterMol

Conversion tool for molecular simulations
MIT License
187 stars 54 forks source link

Multiple entries for same molecule type in [ molecules ] break G->L conversion #310

Open ctk3b opened 7 years ago

ctk3b commented 7 years ago

[ system ] directives that contain multiple entries for the same molecule type, e.g.,

SOL 10
SOL  5

end up writing incorrect bond/angle/dihedral indices in the LAMMPS output.

mrshirts commented 7 years ago

Any immediate thoughts for what the basic issue is? There are two ways to do it: treat them as literally different molecules, or try to merge them. Errors like this would occur if what is being done is halfway in between.

ctk3b commented 7 years ago

Yeah it's an accounting error somewhere involving the offset variable. Will try to pin it down later today - was mostly documenting this as it was a user reported error via email.

mrshirts commented 7 years ago

Great, sounds good.

dspoel commented 7 years ago

Merging might be tricky when the order it alternating, e.g. SOL 10 Protein 1 SOL 5 Na+ 3 Cl - 5

mrshirts commented 7 years ago

Is this holding you guys up right now or is there a workaround? We are working on prioritizing the existing issues!