[gmx-users] About pdb2gmx and cyclic peptides

Justin Lemkul jalemkul at vt.edu
Thu Jun 26 13:51:27 CEST 2014

On 6/26/14, 6:21 AM, Ivan Gladich wrote:
> Dear all,
>     I am trying to build a cyclic peptide of 12 aminoacids using pdb2gmx.
> Searching on the web I found that is not very straightforward without editing
> the topology and the .gro file
> https://mailman-1.sys.kth.se/pipermail/gromacs.org_gmx-users/2012-September/074609.html

No need to go all the way back to 2012; this issue was discussed at length just 
last week :)

> The problem is that I have thousand of structures and so I need to do it
> iteratively.
> I am wondering if in the newer versions of gromacs the problem has been
> addressed and solved.

pdb2gmx was not designed with cyclic peptides in mind, but it can be made to 
work.  If you create an entry in specbond.dat to link the N-terminal N and 
C-terminal C, while selecting "None" for the respective termini (and using 
-missing to override the "dangling bonds" warning), then it should work, 
assuming the following: (1) you are not using an AMBER force field, because they 
use fixed N- and C-terminal residues that cannot be modified and (2) you have 
all atoms (including H) built onto the starting structure (because -missing 
creates all sorts of problems, and is used here only as a complete workaround).

Whether or not that's feasible for you to do for thousands of peptides (since N- 
and C-terminal residue names have to be set in specbond.dat, and since you need 
to have fully protonated structures as input) is up to you.



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