^ agree with Carlos. You'll have to modify the code of this is strictly
what you need, and you will get very large force discontinuities. One
thing you might consider is simply zeroing out the LENNARD_JONES_BCOEF
values for the interactions you want to have no attractive potential. That
way the r12 term will stay in place, which is mostly what determines the
location of the wall, and there won't be a problem with discontinuous
forces or sharp turns that limit your timestep.
Dave
On Fri, Jul 29, 2016 at 11:46 AM, Carlos Simmerling <
carlos.simmerling.gmail.com> wrote:
> I don't know if there is a built in function for this in Amber, likely you
> will need to modify the code. One thing I wanted to point out though is
> that the forces are quite high at that first crossing, so you'd probably
> need a smoothing function as well to keep the simulations stable. If you do
> as you suggested and just truncate the potential, you'll have severe force
> discontinuities.
> Carlos
>
> On Jul 29, 2016 11:38 AM, "Irem Altan" <irem.altan.duke.edu> wrote:
>
> > Hi,
> >
> > Thank you for your response. I’d like to zero out the potential beyond
> the
> > first point the potential crosses the r-axis. Therefore there will only
> be
> > a repulsive core remaining. This has to be specific for each atom type,
> as
> > the point that crosses the axis is specific to the atom. Would LJEDIT be
> > able to change the form of the potential function like this? If so, where
> > can get more information as to how to use LJEDIT?
> >
> > Best,
> > Irem
> >
> > > On Jul 29, 2016, at 10:55 AM, David Cerutti <dscerutti.gmail.com>
> wrote:
> > >
> > > In what way would you like to "truncate" the interaction? It is
> possible
> > > to specify NBFIX-like terms in Amber (LJEDIT), either as a new section
> of
> > > the parameter file (the thing that's called in order to build
> topologies)
> > > or, I think, using ParmEd. These LJEDIT terms are pair-specific
> > > Lennard-Jones interactions, so you could modify the interaction for
> > > particular types of atoms interacting with one another. The behavior
> is
> > to
> > > leave the standard combining rules in effect unless otherwise stated,
> so
> > > all other interactions between said particles would stay the same.
> > >
> > > If the particles are few in number and always bonded to one another in
> > > fairly close proximity, you might be able to get away with some fancy
> NMR
> > > restraints that have no effect for r > L, where L is something like 4-8
> > A,
> > > and for r <= L the restraints attenuate or cancel out the effect of the
> > > Lennard-Jones terms you want to get rid of. I'd suggest trying LJEDIT
> > > first.
> > >
> > > Dave
> > >
> > >
> > > On Fri, Jul 29, 2016 at 9:52 AM, Irem Altan <irem.altan.duke.edu>
> wrote:
> > >
> > >> Hi,
> > >>
> > >> I was wondering, is it possible to use modified potentials in Amber
> for
> > >> nonbonded interactions? Specifically, I would like to truncate the LJ
> > >> potentials at the repulsive core for some protein atoms. Those atoms
> > would
> > >> interact with the truncated potential amongst themselves, but use the
> > full
> > >> form of the potential when interacting with other atoms. Is it
> possible
> > to
> > >> do this with the tools available in Amber?
> > >>
> > >> Best,
> > >> Irem
> > >> _______________________________________________
> > >> AMBER mailing list
> > >> AMBER.ambermd.org
> > >>
> >
> https://urldefense.proofpoint.com/v2/url?u=http-3A__lists.ambermd.org_mailman_listinfo_amber&d=CwICAg&c=imBPVzF25OnBgGmVOlcsiEgHoG1i6YHLR0Sj_gZ4adc&r=r1Wl_e-3DAvYeqhtCRi2Mbok8HBpo_RH4ll0E7Hffr4&m=uT0dRBXBrATb4KWBlEygpIaNB1pSCb9up3bi7imoE1A&s=EyqUK30O0g83tmSqbVX5T3SYYi_Pbbi8GLcYm7pnjac&e=
> > >>
> > > _______________________________________________
> > > AMBER mailing list
> > > AMBER.ambermd.org
> > >
> >
> https://urldefense.proofpoint.com/v2/url?u=http-3A__lists.ambermd.org_mailman_listinfo_amber&d=CwICAg&c=imBPVzF25OnBgGmVOlcsiEgHoG1i6YHLR0Sj_gZ4adc&r=r1Wl_e-3DAvYeqhtCRi2Mbok8HBpo_RH4ll0E7Hffr4&m=uT0dRBXBrATb4KWBlEygpIaNB1pSCb9up3bi7imoE1A&s=EyqUK30O0g83tmSqbVX5T3SYYi_Pbbi8GLcYm7pnjac&e=
> >
> > _______________________________________________
> > AMBER mailing list
> > AMBER.ambermd.org
> > http://lists.ambermd.org/mailman/listinfo/amber
> >
> _______________________________________________
> AMBER mailing list
> AMBER.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber
>
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Fri Jul 29 2016 - 10:00:02 PDT
