[gmx-users] .top file after using amber2gromacs

Justin Lemkul jalemkul at vt.edu
Mon Mar 3 05:34:11 CET 2014

On 3/2/14, 11:24 PM, Chetan Mahajan wrote:
> Hi All
> I have gromacs files generated using acpype tool acting on Amber files
> originally.
> I need to know the meaning of certain terms in that *.top file (email with
> attachment of .top file got blocked due to attachment). Unfortunately, I
> can't find on the web, any material regarding that. It would be great, if
> anyone could comment on the following.
> *1. What is the meaning of terms under defaults option at the top of the
> file (pasted below)?*
> [ defaults ]
> ; nbfunc        comb-rule       gen-pairs       fudgeLJ   fudgeQQ
>       1                    2                 yes                  0.5
>   0.8333

Please refer to manual section 5.7.1, where each of these terms is explained 
after the urea.top example.

> * 2a. ** Is atomtypes section needed? (sample .top file on gromacs website
> does not have it) (pasted after 2b question)**In this atomtypes section,
> the Amber2gromacs tool keeps mass and charge of each atom as zero. Would a
> simulation with these places zero be a valid simulation **?* My guess is
> yes. THis is since, corresponding values are available later in atoms
> section. The purpose of asking this question is that I have to be sure of
> the runs I have made, that nothing has gone wrong in having those zeros.

Charge and mass information in [atomtypes] is indeed over-written by whatever is 
found in [atoms].  You can always confirm what has been used by obtaining the 
grompp-processed topology with grompp -pp or by using gmxdump on the .tpr file.

> * 2b.I see several enigmatic terms such as bond type p type, Amb. Can you
> explain these terms, if they are necessary at all? Why last two columns for
> Amb?*

Different force fields work in different ways, so the Gromacs file format is 
standard across all the different force fields.  Since there are generally fewer 
types of bonded interactions, bonded types are a subset of nonbonded types. 
Sometimes there are no differences, as is the case here.  If a force field uses 
separate bonded and nonbonded types, that just means the bonded types are an 
interpretation of the atom types used within ffbonded.itp.  The ptype column is 
particle type - A for atoms, S for shells, V for virtual sites.  The last two 
columns are a comment, likely the original AMBER parameters so you can verify 
the unit conversion.

> (part of the data for both 2a and 2b pasted below)
> [ atomtypes ]
> ;name   bond_type     mass     charge   ptype   sigma         epsilon
> Amb
>       Ti       Ti              0.00000  0.00000   A     1.39461e-01
> 6.08772e-02 ; 0.78  0.0145
>      OT       OT           0.00000  0.00000   A     2.87832e-01
> 8.29687e-02 ; 1.62  0.0198
>      HW       HW          0.00000  0.00000   A     0.00000e+00   0.00000e+00
> ; 0.00  0.0000
> *4.WHat is cgnr? how is it different from nr? (it appears in atoms section,
> example below)*

cgnr = charge group number
nr = atom number


>   nr  type  resi  res  atom  cgnr     charge      mass       ; qtot
> bond_type
>       1   Ti     1   iO2    TI    1     1.691002     47.86700 ; qtot 1.691
> Thanks


Justin A. Lemkul, Ph.D.
Ruth L. Kirschstein NRSA Postdoctoral Fellow

Department of Pharmaceutical Sciences
School of Pharmacy
Health Sciences Facility II, Room 601
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