[gmx-users] Constraining starting configuration in Coarse-Grain Simulation

Maghesree Chakraborty maghesree.c at gmail.com
Mon Aug 7 01:24:10 CEST 2017


I am trying to run a CG simulation with only non-bonded interactions. I
have obtained the tabulated potentials for those interactions by
force-matching using VOTCA. I want all my bonds and angles to be
constrained as they are in the initial cg_conf.gro file. Since I do not
have tabulated potentials for bonded interactions, gromacs gives error if I
do not provide the bond and angle parameters (b0,kb and θ0 , kθ) in the
topology file. Then I tried setting the bond parameter (b0) to be the mean
of bond-length distribution in cg_conf.gro and the angle parameter (θ0) to
be the mean of the angle distribution in cg_conf.gro, I constrain the bonds
and angles with LINCS in my grompp file. During simulation, I get a lot of
LINCS warning and simulation is aborted. I am wondering if there is a way
of constraining the initial configurations of the molecules as they are in
cg_conf.gro file  in the absence of tabulated bonded potentials and not
forcing the molecules to have the bond-lengths and angles as specified  by
the topology parameters.

Thank you.

M. Chakraborty

More information about the gromacs.org_gmx-users mailing list