[gmx-users] Can we convert a triclinic simulation box to orthogonal box in GROMACS without loosing periodicity?

Lakshman Ji Verma lakshmanjivrm at gmail.com
Mon Apr 29 22:08:32 CEST 2019

Dear GROMACS users,

I am trying to simulate a triclinic crystal at the center with solvent on
either side in the z-direction and periodic x-y plane. The crystal box is
very skewed with one of the angles from the horizontal axis is 30 degrees.
I have first simulated it in the triclinic box but apparently, a box bigger
in z-direction required. In doing so, I would need much larger x-y
dimensions to satisfy requirements for the triclinic box ( ax>2bx, ax>2cx,
by>2cy). This will increase the cost of simulation significantly.

I tried using trjconv command by centering the molecule around an atom. The
wrapped molecules turn out to be in the orthogonal coordinate system but
the box is still triclinic. I tried removing the non-diagonal elements of
the box but then crystal loses its periodicity.

Is there a way I can convert this triclinic box to orthogonal without
losing the periodicity of the crystal?

My crystal is consist of molecules consist of 74 atoms. Please let me know
if more information is needed.


More information about the gromacs.org_gmx-users mailing list