[gmx-users] Custom urea topology

[Gabriele Lanaro]

Hi! I'm trying to write a custom topology for the urea molecule. Taken from
the following paper http://pubs.acs.org/doi/abs/10.1021/jp030534x.

I generated the bonds, angles, torsional dihedrals and improper dihedrals.
The scenario is as follows:

1) If I leave the bonds section the thing works with or without constraints
2) If I include the angles and leave constraints off. It works. If I put
constraints on, it gives segfault due to lincs errors.
3) If I include the dihedrals it doesn't work, with or without LINCS

The simulation consists in a single urea molecule, I used a really small
timestep to figure out what's happening.

I'm having hard time spotting where the problem exactly is and how to solve
it. Any help would be appreciated.

Gabriele

PS. I could try to use other forcefields (like amber) and start from that,
but I'd like to understand how the gromacs topologies work, so that I can
write a topology generator and additional documentation.