[gmx-users] How to generate opls aa parameters for new element myself?

[Clinton King]

The procedure as I understand is that you need to first edit the OPLSAA
files. I found that it worked best if I created a copy of the whole folder.
These files are in your installation directory. You'll need to edit the
atomname2type.n2t file at a minimum. You'll need to create a type for the
Pt atoms. This will require you to add the mass and partial atomic charges.
If the Pt is in a complex, you'll want to get the partial atomic charge
from a QM computation using something like Gaussian 09 or NWChem. If the Pt
is an individual atom, you can try setting it to zero and see if that
works. If the Pt is heterogeneous, then I don't know whay you'll need to
do, as I've never done mm simulation of heterogeneous systems. If the Pt is
in a complex, you'll also need to put bond lengths in.
You'll need to edit the force field potentials. I don't remember off the
top of my head what those file names are, but they are in the same folder
as atomname2type.n2t. You'll need to put in Lennard-Jones parameters for
non-bonded interactions, which you may have to guess at at first. If the Pt
atom is in a complex, you'll need to add bonded interaction parameters,
angle bending parameters, and dihedral parameters. Some (or all) of those
can be obtained from a QM frequency calculation, but it's not very
straightforward, and I have not successfully used that route.
You'll then have to use the x2top utility to generate your topology file.
It will probably take several tries before you get everything right. Once
you have a topology file in hand, you should be able to generate your .tpr
file and then start a molecular dynamics run.
You'll want to benchmark and optimize the parameters by generating some
known experimental value (you'll have to dig into the literature, maybe you
can find solvated structures or a radial distribution function) and then
adjusting the parameters until the experimental value is reproduced. You
could also benchmark the molecular mechanics parameters against a
higher-level qm computation by seeing if they accurately reproduce qm bond
lengths, etc.

--
Clinton King
Graduate Student
Chemistry Department
Brigham Young University


>
> Message: 4
> Date: Tue, 31 May 2016 04:51:54 +0000
> From: Wang Tao <wangtaoone at hotmail.com>
> To: "gromacs.org_gmx-users at maillist.sys.kth.se"
>         <gromacs.org_gmx-users at maillist.sys.kth.se>
> Subject: [gmx-users] How to generate opls aa parameters for new
>         element myself?
> Message-ID:
>         <
> SN1PR07MB2398F47ED8CA4876AFAB36ABCE460 at SN1PR07MB2398.namprd07.prod.outlook.com
> >
>
> Content-Type: text/plain; charset="iso-8859-1"
>
> Hello,everyone!
>
> I'm to conduct a MD simulation of a system in which there are plenty of
> neutral Pt atoms . My software is gromacs, and the foece field is opls aa.
>
> The proplem is ,there is no avalable opls aa force filed parameters for
> Pt.So I plan to generate one myself.But i don't know what the correct
> process is .Can any one provide any instructions,please?
>
> Thank You
>
> Tao Wang
>
>
>
>