[gmx-users] Charmm 36 parameter file for gromacs

Justin Lemkul jalemkul at vt.edu
Wed Aug 12 13:33:06 CEST 2015

On 8/11/15 10:15 PM, Vy Phan wrote:
> Dear Dr. Justin and Gromacs User,
> Thank you so much for your helps.
> When I use charmm36 ff to generate the topology for lipid (DOPC). I got
> many warning "Warning :Long bond "
> And after get the topology, the total charge is very high (+888 e for DOPC
> bilayer have 516 molecular). When I used grompp to generate the .tpr file,
> I got the many error like this  :
> Error  3379 [file topol_Other_chain_D.itp, line 198252]
>   No defaut U-B types
> ...
> Error  2379 [file topol_Other_chain_D.itp, line 98252]
>    No default Proper Dih.type
> ....
> What can I do to modified the topology to get the correct charge and  fix
> the error ?
> I download dopc.itp from Tieleman's website. I run simulation and all
> things look fine.  People have experiences, please tell  me the difference
> of this topology and the one on the charmm36 force field.

The files Tieleman provides are united-atom structures that may use a different 
naming convention.  They won't have H atoms on the aliphatic groups and there is 
no .hdb entry for DOPC.  The ridiculous net charge you got for DOPC (which is 
neutral!) should be an indicator that whatever topology you built is utter 
nonsense.  There should also be no warnings about missing parameters, so this is 
further evidence that the topology is wrong and the "things look fine" assertion 
seems totally unfounded to me.

If you want to simulate a simple membrane, use CHARMM-GUI to build it.  It will 
be 100% CHARMM compliant and we give you all the necessary input files.



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