[gmx-users] GROMPP replaces lipid atom coordinates with these of water molecules

[Justin Lemkul]

On 7/23/12 9:54 AM, Christian Blouin wrote:
> Hello,
>
>         I have been running a simple simulation where I replace 150
> water molecules in a waterbox with a lipid called DPC to create
> micelles. This simulation is working fine and I'm getting decent
> micelles fairly quickly. I am now trying to do the same, but with a
> small peptide in the simulation box. I think that grompp is using
> water molecules coordinates to locate the DPC, making non-sensical
> molecules. Why is grompp doing this in the simulation with the peptide
> and not without is what stomps me. Here follow more details about my
> system and errors:
>
> A PDB of this atomic mess can be found here:
> http://www.cs.dal.ca/~cblouin/gromacs/gromppout.pdb
>
> I've been working on this for a while now, with no success. Can anyone
> point me in the right direction?
>
> Thanks,
>
> Christian
>
> --------------------------------------------------
>
> Here are the command lines that I use:
>
> pdb2gmx -f 2LOU.pdb -o protein.gro -i ./include.itp -p protein.top -ignh
> editconf -f protein.gro -o protein-PBC.gro -bt dodecahedron -d 1.0
> genbox -cp protein-PBC.gro -cs spc216.gro -ci dpc -nmol 150 -p
> protein.top -o protein-water.gro
> python edittop.py (this python script adjust the protein.top file for
> the right count of DPC and adds the include statement for DPC).
>
> grompp -f minim.mdp -c protein-water.gro -p protein.top -o
> protein-water.tpr -maxwarn 4
>
> The last grompp command generates the following warning:
> Warning: atom name 665 in protein.top and protein-water.gro does not
> match (OW - C1)
> Warning: atom name 666 in protein.top and protein-water.gro does not
> match (HW1 - C2)
> Warning: atom name 667 in protein.top and protein-water.gro does not
> match (HW2 - C3)
> ...
> WARNING 1 [file protein.top, line 4194]:
>    6900 non-matching atom names
>    atom names from protein.top will be used
>    atom names from protein-water.gro will be ignored
>

This is the problem.  The order of the [molecules] directive must match that of 
the coordinate file.  If it does not, grompp does precisely what you're seeing - 
it maps coordinate and topology information incorrectly, causing the 
rearrangement you're seeing.

-Justin

> If I extract a PDB from the tpr file, the system looks fine except for
> all of the DPC atoms are now crammed into a rectangular box at the
> edge of my PBC. In fact, it looks like the coordinates used for the
> DPC molecules are these of water molecules that got reassigned to the
> lipids residues.
>
> The topology and gro file (if this helps):
> http://www.cs.dal.ca/~cblouin/gromacs/protein.top
> http://www.cs.dal.ca/~cblouin/gromacs/protein-water.gro
>

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

Justin A. Lemkul, Ph.D.
Research Scientist
Department of Biochemistry
Virginia Tech
Blacksburg, VA
jalemkul[at]vt.edu | (540) 231-9080
http://www.bevanlab.biochem.vt.edu/Pages/Personal/justin

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