[gmx-users] Trouble running simulations with Nitrogen

Justin Lemkul jalemkul at vt.edu
Fri Jul 20 14:14:30 CEST 2018

On 7/19/18 11:54 PM, Joshua Cummings wrote:
> Hi Everyone,
> I'm currently trying to run some simulations that involve the absorption of N2 into a liquid solvent. I'm having some difficulty in generating the topology for N2. For reference, the pdb file I have for N2 is as follows:
> HETATM    1  N6B HDZ     0       0.000   0.000   0.553  1.00  0.00           N
> HETATM    2  N6A HDZ     0       0.000   0.000  -0.553  1.00  0.00           N
> CONECT    1    2
> CONECT    2    1
> When I try to generate the .itp file and topology via pdb2gmx (OPLS), gromacs doesn't recognise the residue. I've tried using both the LigParGen server and the PRODRG to generate the topology, but neither work. PRODRG generates an error saying it does not deal with mono/di-atomic molecules and LigParGen won't recognise N2 at all.
> I haven't had any trouble generating topologies for other small gas molecules such as CH4, am I missing something here or are OPLS parameters not possible for N2? Should I use some other forcefield for N2?

There should be parameters in the literature for N2, perhaps not for 
OPLS-AA (the "L" is for "liquid," after all, and focuses on biomolecules 
and analogs) but might give you a place to start. In principle, the 
parametrization is pretty straightforward. The charges are zero, the 
bond length and vibrational frequency can be easily computed via QM if 
not determined experimentally via spectroscopy, and then you need to fit 
the LJ parameters to reproduce either liquid density/RDF at low 
temperature, free energy of solvation (if known), etc.



Justin A. Lemkul, Ph.D.
Assistant Professor
Virginia Tech Department of Biochemistry

303 Engel Hall
340 West Campus Dr.
Blacksburg, VA 24061

jalemkul at vt.edu | (540) 231-3129


More information about the gromacs.org_gmx-users mailing list