[gmx-users] Polymer in periodic boundary conditions again

[Ester Chiessi]

Hi all,

I asked my question about polymer in pbc recently, with several interesting
answers.

K.A. Feenstra wrote:

> Ester Chiessi wrote:
> >
> > Hi gromacs users,
> >
> > I'm trying to build an infinite chain in periodic boundary conditions..
> >
> > I saw I have to use:
> >
> > setenv GMXFULLPBC.
> >
> > The question is how to insert in the topology file the information that
> > the last atom of the chain has to be covalently bound to the periodic
> > image of the first .
>
> There is no distinction between an atom and any of its periodic images.
> So, you simply define a bond between the appropriate atom in the last
> residue and the atom in the first residue.
>
> > Should I have to use the same residue name for each repetitive unit (
> > i.e. no end-residues) only giving the connectivity to the preceeding
> > residue and then choose the exact box lenght to contain the part of the
> > chain I'm defining in the topology?
>
> Yes.
>
> > Is it important to start with an orientation of the chain which
> > geometrically satisfies the periodicity requirement?
>
> Yes.
> You can use editconf to orient your polymer along its principal axes.
> That will probably already be very close to the periodically
> consistent orientation.
>
> > Is it possible to solvate the polymer and use the PBC jet?
>
> Yes.

Now I'd need help after some unsuccessful trials.

I begin with a simple system to set the method.

I built a cubic box ( 2.576 nm ) of SPC waters with a linear polymer chain
centered in and aligned along the X axis. The first atom of the polymer is
at x=0.033 and the last at 2.476 nm, in the periodically consistent
orientation to be bound to the image of the first atom.
I explicitly added in the top file the bond between the last and the first
chain atom and angles, dihedrals and pairs to complete the connectivity.
Then I started a dynamic run (timestep 0.002 ps), which crashed.
>From md0.log:
         Removing pbc first time
         There were 2 inconsistent shifts. Check your topology.
 .........
          Constraint error in algorithm Lincs at step -2
 .........
          Constraint error in algorithm Lincs at step -1
 .........
          Constraint error in algorithm Lincs at step 0
..........
          t=0.000 ps. Water molecule starting at atom 389 can not be
settled. Check for bad contacts and/or reduce the timestep.

In another test, I repeated the same calculation but without introducing the
connectivity in the topology file. The run chrashed again, with a constraint
error in algorithm Lincs at step 1. In the step0.pdb output file some atoms
of the chain ends are far off the box.

Then I tried the following (with a topology without explicit connectivity).
Before starting the dynamics, I minimized the energy of the system to relax
the water, by constraining the polymer chain at the aim to mantain the
orientation for the periodicity. After this calculation the chain ends
diverged a little. The following dynamics seems to go, but the chain ends
move without periodicity.

In each trial I used: setenv GMXFULLPBC 1.

Where is the mistake?

Thanks a lot for any suggestion!

Ester


--
Ester Chiessi
Dipartimento di Scienze e Tecnologie Chimiche
Universita' di Roma "Tor Vergata"
Via della Ricerca Scientifica
00133 Roma (Italy)
http://www.stc.uniroma2.it/cfmacro/cfmacroindex.htm
e-mail: ester.chiessi at uniroma2.it
Phone: 39*6*72594462
       39*6*72594874
Fax:39*6*72594328