[gmx-users] adding ATP parms to AMBER99

[Justin Lemkul]

On 4/11/14, 1:43 PM, Stephen N. Floor wrote:
> Hi gromacs users -
>
> I am trying to simulate a protein-ATP complex and am trying to add parameters
> for ATP to the forcefield amber99sb-ildn.  First off, I find it likely that
> someone has already done this - is there an .itp file for ATP floating around
> somewhere with the recommended parameters from Meagher, Redman & Carlson, JCC
> (2003)?  http://www.ncbi.nlm.nih.gov/pubmed/12759902
>

The .rtp and .hdb files that you would need to create the topology with pdb2gmx 
appear to be available from:

http://bugzilla.gromacs.org/issues/721

You still need to modify ffbonded.itp and ffnonbonded.itp to use the published 
parameters, since they differ from those already in the force field.

> If not, I am trying to decide between using parms generated from PRODRG
> (which spits out a .itp, albeit with non-amber atom types) versus editing the
> ffbonded and ffnonbonded files for the forcefield to include bond, angle, and
> dihedral parms from Meagher 2003.  Any thoughts on this?  I am partial to
> using parameters from Meagher 2003 since they did a full QM treatment of the
> ligand and these are recommended by Bryce.
>

Abort any notion of using PRODRG.  The topology is of no use to you, because you 
literally have to change everything about it.

> I tried to follow previous messages in the gmx-users list on this topic, but
> am stuck translating dihedrals from the AMBER94 FRCMOD file to ffbonded.itp,
> as recommended in a previous thread
> (http://gromacs.5086.x6.nabble.com/Re-AMBER-force-fields-for-ATP-td4440113.html)

If you can explain where you're stuck, probably someone can suggest something to 
help.  Without knowing what "stuck" means, there's not much to go on :)

> For the AMBER94 PREP and FRCMOD files please see here:
> http://www.pharmacy.manchester.ac.uk/bryce/amber/
>
> I think I could generate input files that would be properly parsed, but am
> looking for verification that I'm doing it correctly before conducting
> simulations with incorrect parameters.
>

All that's necessary is translation between file formats (and units); there are 
many scripts online floating around that do this.  Your approach (aside from the 
PRODRG tangent) is correct.

The simplest test to convince yourself that everything is OK is to do a 
single-point energy evaluation within Gromacs and then in AMBER; if the energies 
are the same, the force fields are equivalent.

-Justin

-- 
==================================================

Justin A. Lemkul, Ph.D.
Ruth L. Kirschstein NRSA Postdoctoral Fellow

Department of Pharmaceutical Sciences
School of Pharmacy
Health Sciences Facility II, Room 601
University of Maryland, Baltimore
20 Penn St.
Baltimore, MD 21201

jalemkul at outerbanks.umaryland.edu | (410) 706-7441
http://mackerell.umaryland.edu/~jalemkul

==================================================