[gmx-users] Getting a crystal topology file from pdb2gmx

[Justin Lemkul]

On 11/11/15 12:02 PM, Ben Tam wrote:
> Dear Justin,
>
> Thank you for your prompt reply Justin. I think that is what I was looking for but did not know such command it existed! I have tried your suggestion however another problem occurs :
> Opening force field file /shared/ucl/apps/gromacs/5.1.1/intel-2015-update2/share/gromacs/top/oplsaa.ff/atomname2type.n2tThere are 23 name to type translations in file oplsaa.ffGenerating bonds from distances...atom 618Can not find forcefield for atom C6-3 with 3 bondsCan not find forcefield for atom C9-4 with 3 bondsCan not find forcefield for atom C2-7 with 3 bondsCan not find forcefield for atom C1-8 with 3 bondsCan not find forcefield for atom C5-13 with 3 bondsCan not find forcefield for atom C6-20 with 3 bondsCan not find forcefield for atom C9-21 with 3 bondsCan not find forcefield for atom C2-24 with 3 bondsCan not find forcefield for atom C1-25 with 3 bondsCan not find forcefield for atom C5-30 with 3 bonds---------------------------------------------------------------------------------------------------------------------------------------------Can not find forcefield for atom C5-317 with 3 bondsCan not find forcefield for atom N1-322 with 3 bondsCan not find force!
 fie
>   ld for atom C12-323 with 4 bondsCan not find forcefield for atom C11-324 with 3 bondsCan not find forcefield for atom C14-327 with 3 bondsCan not find forcefield for atom C6-346 with 3 bondsCan not find forcefield for atom C9-347 with 3 bondsCan not find forcefield for atom C2-350 with 3 bondsCan not find forcefield for atom C1-351 with 3 bondsCan not find forcefield for atom C5-356 with 3 bondsCan not find forcefield for atom N1-361 with 3 bondsCan not find forcefield for atom C12-362 with 4 bondsCan not find forcefield for atom C11-363 with 3 bonds-------------------------------------------------------------------------------------------------------------------------------------------Can not find forcefield for atom Zr2-590 with 0 bondsCan not find forcefield for atom Zr2-591 with 0 bondsCan not find forcefield for atom Zr2-592 with 0 bondsCan not find forcefield for atom Zr2-593 with 0 bondsCan not find forcefield for atom Zr2-594 with 0 bondsCan not find forcefield f!
 or a
>   tom O3-595 with 0 bondsCan not find forcefield for atom O3-596 with 0 bondsCan not find forcefield for atom O3-597 with 0 bondsCan not find forcefield for atom O3-598 with 0 bondsCan not find forcefield for atom O3-599 with 0 bondsCan not find forcefield for atom O3-600 with 0 bondsCan not find forcefield for atom O3-601 with 0 bondsCan not find forcefield for atom O3-602 with 0 bondsCan not find forcefield for atom O3-603 with 0 bondsCan not find forcefield for atom O3-604 with 0 bondsCan not find forcefield for atom O3-605 with 0 bondsCan not find forcefield for atom O3-606 with 0 bondsCan not find forcefield for atom O3-607 with 0 bondsCan not find forcefield for atom O3-608 with 0 bondsCan not find forcefield for atom O3-609 with 0 bondsCan not find forcefield for atom O3-610 with 0 bondsCan not find forcefield for atom O3-611 with 0 bondsCan not find forcefield for atom O3-612 with 0 bondsCan not find forcefield for atom O3-613 with 0 bondsCan not find forcefield fo!
 r at
>   om O3-614 with 0 bondsCan not find forcefield for atom O3-615 with 0 bondsCan not find forcefield for atom O3-616 with 0 bondsCan not find forcefield for atom O3-617 with 0 bondsCan not find forcefield for atom O3-618 with 0 bonds
> -------------------------------------------------------Program gmx x2top, VERSION 5.1.1Source code file: /dev/shm/tmp.PwtfSrm0uJ/gromacs-5.1.1/src/gromacs/gmxpreprocess/x2top.c, line: 216
> Fatal error:Could only find a forcefield type for 420 out of 618 atomsFor more information and tips for troubleshooting, please check the GROMACSwebsite at http://www.gromacs.org/Documentation/Errors
>

Sorry, not going to be able to make heads or tails of that mess without proper 
line endings and whatnot.  Something is total garbage in the .n2t file.

> With the co-ordinate I had, I have no problem visualising in gaussian view. So what can I do to fix this? Does that mean I have to write a forcefield and incorporate into GROMACS? I know this sounds very rookie, but what is the format of .n2t and do I put that file in my working directory? Thanks for your help.

Visualization has no implication on whether or not GROMACS can understand the 
contents and generate a topology.  Unfortunately, it appears that the .n2t 
description was also nuked from the website (put that on my to-do list of 
amending the documentation), but look at the example from OPLS, for instance:

C    opls_157    -0.18  12.011  4    H 0.108   H 0.108   H 0.108   C 0.150

The columns are:

1. Atom of interest (element, not the exact name of the atom)
2. Atom type for that atom
3. Charge of the atom
4. Mass of the atom
5. How many bonds it can form
6. The remaining columns are the the elements to which the atom in column 1 is 
bonded, with their expected lengths.

Bonds are only formed if the actual distances in the coordinate file are within 
+/- 10% of the specified lengths.

You should be doing all of this in a custom force field directory, almost certainly.

-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

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