GAFF or Amber99sb-ildn for simulating small molecule binding with protein containing metal ions

User discussions
1

GROMACS version: 2023.3-plumed_2.10.0_dev
GROMACS modification: Yes (with plumed)

Hi,

I am trying to run MD simulations for a small molecule in the binding pocket of a target protein.

I know usually GAFF is usually used for small molecule simulations and is prepared with ANTECHAMBER. However, the target protein also contains a metal ion that is quite important to the protein stability and is also close to the binding pocket. Usually the target protein is simulated with the Amber99sb-ildn. The small molecule is also technically composed of amino acids but many of them are heavily modified and unnatural amino acids. The ligand is also quite small so I’m thinking it may be more easier to just treat it as a small molecule ligand.

What force field should I be using for this situation? What I’m thinking is either:

  1. Parameterize the target protein with Amber99sb-ildn and the small molecule ligand with GAFF. However, I’m not sure if this would be a correct thing to do. Technically the simulation does not need to be super strict, I just want to confirm if there is stable binding.

  2. Try and parameterize the small molecule with Amber99sb-ildn, somehow. I’m not sure if this would even work as I previously tried using Amber99sb-ildn with some modified amino acids before and it was very painful, requiring a lot of custom modifications. The ligand is much more complicated than simple modifications however and I’m not sure if it is even possible. If there is a good way to do this however I would love to know.

  3. Just parameterize everything with GAFF. I don’t favor this option however as the metal ion is quite critical to the stability of the target protein and may also affect the binding energy with the ligand. From my understanding, Amber99sb-ildn can properly account for this, but for GAFF I’m not too sure. If my understanding is incorrect and GAFF is also capable of accounting for metal ions, then I suppose this would be the easiest option.

Thanks in advance for any suggestions.

Edit: I saw this (https://ambermd.org/tutorials/advanced/tutorial20/mcpbpy.php) which seems quite relevant to what I’m trying to do. I am currently running everything in GROMACS although I could switch to a different MDS software if necessary. Would the linked tutorial be the best approach?

Edit 2: As I look more into this, it seems that the GAFF used by Antechamber was designed to be mixable with other AMBER forcefields. Also, the post The use of GAFF1 vs GAFF2 with the pmx ff99SB*-ILDN forcefield - pmx - BioExcel and the mentioned paper: ‘Large scale relative protein ligand binding affinities using non-equilibrium alchemy’ seem to succeed that this is indeed ok. I guess this would be the simplest approach in that case, to parametarize the ligand with Antechamber, use Amber99sb-ildn for the ion containing target protein, and run the workflow in GROMACS. Does this seem reasonable, or is there anything else I need to watchout for?