[gmx-users] No interaction between manually added atoms

[Justin Lemkul]

On 11/15/16 7:44 AM, Kamps, M. wrote:
> Dear GMX-users,
>
> I'm experiencing difficulty with my manually added atoms to create a
> metallic surface within GROMACS. My goal is to create a surface (at this
> point metallic, Al) and let it interact with a protein.
> In my current situation I've only modelled the aluminium surface
> (non-bonded surface). The EM drops to a minimum in only 8 steps (first red
> flag?), the NVT and NPT equilibrium look ok, but during the MD run the
> surface disassembles like it is made up of loose sand. There is no
> interaction at all between my individual atoms. What did I do wrong? Or
> what did I forget to do?
>
> I created the surface by defining one single Aluminium atom in a pdb file,
> which i converted via pdb2gmx. To enable this, I added a Al.rtp file to a
> CHARMM27 forcefield, this file contains these lines:
>
> [ bondedtypes ]
> ; bonds  angles  dihedrals  impropers
>      1       1          1          1
> [ AL ]
>  [ atoms ]
> ; name  type  charge  chargegroup
>     AL    AL   0.000     0
>
> Furthermore, to the atomtypes.atp file I added this line:
>
> AL    26.981539 ;    Aluminium mass
>
> To the ffnonbonded.itp file I added this line:
>
> AL    13    26.981539    0.000    A    0.2617500000    0.300000
>
> Where the last two numbers relate to the Lennard-Jones potential. Although
> these values are not based on any scientific literature (at this point),
> they should show some interaction?
>

Garbage in, garbage out.  If you don't have a basis for knowing that these 
parameters give reasonable behavior, "some interaction" may be exactly what 
you're getting.  Use real parameters, from literature, or otherwise derived 
properly.

For instance, CHARMM-compatible metal parameters for FCC lattice structures can 
be found in:

Hendrik Heinz, R. A. Vaia, B. L. Farmer and R. R. Naik, Accurate Simulation
of Surfaces and Interfaces of Face-Centered Cubic Metals Using 12-6 and 9-6
Lennard-Jones Potentials, Journal of Physical Chemistry C, 2008, 112, 17281-17290.

> When I put the atom into pdb2gmx this is my output Al.gro file:
>
> ALUMINIUM
>     1
>     1AL      AL    1   0.000   0.000   0.000
>    0.00000   0.00000   0.00000
>
> This atom is then replicated with insert-molecules to form an FCC crystal,
> using an xyz coordinate file. The result is a nicely stacked (small slab
> of) surface consisting of 256 atoms. The topology file looks like this:
>
> [ moleculetype ]
> ; Name            nrexcl
> AL               3
>
> [ atoms ]
> ;   nr       type  resnr residue  atom   cgnr     charge       mass
> typeB    chargeB      massB
> ; residue   1 AL  rtp AL   q  0.0
>      1         AL      1     AL     AL      1          0    26.9815   ;
> qtot 0
>
> Including the usual lines such as include water models, ion models etc..
> The surface is then put in a box (with 1.5 nm distance to edges, in all
> directions) and solvated with tip3p water molecules. Can someone help me?
>

If you have water, then you need to know if the assigned LJ parameters produce 
meaningful interactions with water.  Solve one problem at a time; deal with the 
pure metal lattice and make sure its properties are reasonable.  Then figure out 
if those parameters are balanced with respect to water interactions.

-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

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