[gmx-users] Simulation queries

Justin Lemkul jalemkul at vt.edu
Tue Mar 14 13:17:46 CET 2017

On 3/14/17 8:08 AM, Acqualine Lobo wrote:
> Hi,
> I have a couple of questions. I'd be really glad if anybody can help me
> with them.
> 1) The parameters for the Straub Karplus CO model are present in the Charmm
> 36 forcefield, with residue type CO. In order to use that on my ligand
> molecule (CO) I have to change the residue type in my pdb file from CMO to
> CO. That works just fine! But since I have a heme group in my pdb file
> pdb2gmx throws an error as the Charmm 36 forcefield recognizes the heme
> group with residue name HEME and not HEM as it is mentioned in pdb. When I
> alter the pdb file and change HEM t0 HEME the pdb file is disrupted. The
> entire structure is distorted. What should I do?

Edit the file such that you don't shift the coordinate columns.  PDB files are 
fixed-format so you can't shift any columns around.

> 2) Can I copy parameters for CO from Charmm and paste them into
> Gromos53a5/6? Is that even appropriate?

No, this is mixing force fields and is invalid.  Force fields are 
self-consistent entities with different parametrization protocols and levels of 
error.  You can't take a piece of one force field and put it in a different one.

> 3) I actually did the above! Because if I use G53a5/6 pdb2gmx throws an
> error for CO and if I use Charmm pdb2gmx throws an error for HEME. Using
> the above (2) idea I played around with the files and somehow set up a
> simulation. The simulation ran well without any errors but the end result
> was disastrous (as expected). The  CO molecule splits into C and O during
> simulation.

If there's a bond between them, that's not possible, but given that you 
shouldn't be doing this anyway, start over properly using CHARMM.  The only 
difficulty you faced was in renaming HEM -> HEME by (likely) accidentally 
inserting a column when you did so.  If you replace /HEM /HEME/ then you should 
be fine.


> How can I handle this simulation exercise? Which forcefield should I use? I
> want to simulate a Hemoglobin molecule in complex with Carbon Monoxide.
> Thanks in advance for any help!
> Acqualine Lobo


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


More information about the gromacs.org_gmx-users mailing list