Force matching with prmtop files

Bartholomeo Pietzu bartho... at gmail.com
Thu Nov 30 08:44:16 UTC 2017


 I found a neat solution to the problem myself.

Just define a Forcefield ON TOP of your topology. e.g.:

    &FORCEFIELD
      parm_file_name  ./d_mediated_basepair_small_subsys.prmtop
      parmtype AMBER
      &CHARGE
        ATOM Hg2
        CHARGE ${PRM03}
      &END
      &NONBONDED
        &LENNARD-JONES
          atoms Hg2 Hg2
          EPSILON ${PRM01}
          SIGMA [angstrom] ${PRM02}
          RCUT [angstrom] 12.0
        &END LENNARD-JONES
      &END NONBONDED
      &SPLINE
         EMAX_SPLINE [hartree] 10000.0
      &END SPLINE
    &END FORCEFIELD


CP2K will report it found doubly defined FF parameters, but it will use the
manually defined ones. Now you can do a normal force matching.

Greetings


#Dear CP2K Team & Users,
#
#I am trying to run a force matching using amber topology files (prmtop).
#I already ran BO-MD to get reference energies and forces.
#
#I replaced the properties to fit in the prmtop file to match with the
#corresponding variable e.g. ${PRM01} and ${PRM02}.
#
#For the file structure I used the one from tests/optimize_input
#
#now the FITIT-worker-0.out tells me, cp2k does not understand the
variables in #the topology file.
#
#
#
*******************************************************************************

# * ___ *
# *  / \ *
# * [ABORT] *
# *  \___/                 Variable ${PRM01} not defined in
file:               * #
 *    |             ./d_mediated_basepair_matching.prmtop Line:
32         *
# * O/| *
# * /| | *
# * / \ input/cp_parser_inpp_methods.F:466 *
#
*******************************************************************************

#
#
# ===== Routine Calling Stack =====
#
#            9 parser_read_line_low
#            8 parser_read_line
#            7 rdparm_amber_8
#            6 read_connectivity_amber
#            5 connectivity_control
#            4 topology_control
#            3 fist_init
#            2 fist_create_force_env
#            1 CP2K
#
#So I am assuming cp2k can not do a force matching with amber topologies?
Do I #really need to put my amber FF by hand into cp2k?
#
#Greetings
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20171130/9c281b61/attachment.htm>


More information about the CP2K-user mailing list