[gmx-users] acetaminophen charges

[Justin Lemkul]

On 7/4/16 5:19 AM, elham tazikeh wrote:
> Dear GMX usere
>
> my problems is in the assigning the ligand charges in protein-ligand
> simulation
>
> (for instance :acetaminophen )
>
> i acquired (*.itp ) file by PRODRG (after optimization by gaussian
> software) and ATB server as below.
>
> which one is appropriate to my MD simulation by gromaces:
>

PRODRG topologies are extremely unreliable, even for the simplest molecules:

http://pubs.acs.org/doi/abs/10.1021/ci100335w

Don't use PRODRG without significant refinement.  ATB is generally better.

Acetaminophen is a straightforward molecule to parametrize.  GROMOS 
parametrization (which I have described at length in several messages to this 
forum in recent days) is done empirically.  First, identify groups that have 
known parameters and apply them to your molecule.  For acetaminophen, this is 
pretty much everything - aromatic CH groups, aromatic hydroxyl (COH), and an 
amide.  You can generate an initial topology completely from known groups, so 
you should probably start there.  There may be some work needed in terms of 
bonded parameters like dihedrals, but again, start from what is known in the 
force field.

Any topology, no matter its source (manual generation or some server), needs 
validation, so you have to find suitable experimental properties that can be 
tested before using this topology for any production simulation.

>
> *atom*
>
> *By prodrg server*
>
> *By ATB server*
>
> CH3
>
> 0.013
>
> -0.062
>
> C
>
> 0.590
>
> 0.734
>
> O
>
> -0.493
>
> -0.643
>
> N
>
> -0.716
>
> -0.706
>
> H
>
> 0.328
>
> 0.422
>
> C
>
> 0.336
>
> 0.337
>
> CR1
>
> -0.150
>
> -0.229
>
> HC
>
> 0.171
>
> 0.176
>
> CR1
>
> -0.181
>
> -0.286
>
> HC
>
> 0.133
>
> 0.181
>
> C
>
> 0.378
>
> 0.379
>
> OA
>
> -0.650
>
> -0.590
>
> H
>
> 0.405
>
> 0.445
>
> CR1
>
> -0.213
>
> -0.286
>
> HC
>
> 0.118
>
> 0.181
>
> CR1
>
> -0.181
>
> -0.229
>
> HC
>
> 0.124
>
> 0.176
>
> in addition to, in ATB server, i saw this message in the page
>
> (molid; 29980 or 26014):
>
>
> Warning! This molecule contains non-standard atom types not included in the
> standard GROMOS 53A6 and 54A7 forcefield. To use these atom types the
> internal GROMACS parameter files must be updated. These can be downloaded
> using the link below. Instructions on where to place (and how to use) the
> files are provided in the README file included in the archive.
>
>
> Note: my force field was; gromos53A6
>
> what is the meaning?
>

Use the force field files they supply, as there are apparently new atom types 
that are used.  Or follow my suggestions above and build your molecule from 
completely known entities.

-Justin

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

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

Department of Pharmaceutical Sciences
School of Pharmacy
Health Sciences Facility II, Room 629
University of Maryland, Baltimore
20 Penn St.
Baltimore, MD 21201

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

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