GULP and general force fields as Dreading, UFF, GAFF and OPLS-AA

GULP
1

Hi

I am not 100% sure how to correctly generate input files for the complex universal force fields like Dreading, UFF, GAFF and OPLS-AA. I had made all of them working in GULP but only for Dreading and UFF I am sure my procedure is OK (because I can do reference calculation in Forcite+ inside Materials Studio).

My procedure is:

  1. Define the atoms in asymmetric unit cell part

  2. Give unit cell and space group

  3. Manually identify atoms types based on the used force filed.

There do not exist automatic methods for this part - I need to encode them eventually myself …
As far as I know the only one free engine for this work is OpenBabel (it can for sure
generate Dreading atom types - the bgf output) …

  1. Optionally supply atom charges (e.g from Orca or similar QM code) .

  2. Let GULP automatically assign bonds. It is manually possible only in P1 space group, because GULP rearrange the input atoms breaking the already pre-created bond information.

I am not sure whatever:

A) The order of the automatically created bond is always 1. Is it OK for the listed force fields? Do they determine the forces constants from the atom type not form bond type?

B) Are bond angles forces generated automatically OK as well?

C) Are bond torsion angles generated automatically OK as well?

D) Is there any chance to generate library field for MMFF94 force field ?

Let me know, please, if you have any related ideas.

2

I had already find some answers myself:

  1. Bond order 1 is OK for GAFF,Dreading. UFF require correct bond order
  2. I had managed to implement an atom type generation engine based on Antechamber routines
    from the AMBER code
  3. Angles and torsion angles forces are generated OK when defined correctly in force field