Dear David,
Thanks for your suggestions, that would be very helpful!
Just now I read some the UFF related papers and it seems the LJ form for
that FF has an extra parameter, so it cannot be employed in amber directly,
I think.
I'll first try to use MCPB.
2016-06-29 17:11 GMT+08:00 David Cerutti <dscerutti.gmail.com>:
> Oops... edit... be wary of "0.0 and 0.0" for Pt parameters. I don't know
> specifically what UFF has.
>
> On Wed, Jun 29, 2016 at 5:11 AM, David Cerutti <dscerutti.gmail.com>
> wrote:
>
> > I'd be wary of those specific parameters, especially if there's going to
> > be a charge on the Pt. With no Lennard-Jones properties it would be
> like a
> > polar hydrogen, possibly with a very large charge, and significantly more
> > spacing between it and any ligands with steric properties to help shield
> it
> > from bad interactions (i.e. water oxygen comes in and becomes
> > electrostatically attracted to the bare Pt core).
> >
> > You might try putting some nominal Lennard-Jones parameters on the Pt,
> and
> > making use of Pengei Lee's MCPB.py tool (metal center parameter builder).
> > It implements the Seminario method, in particular, for deriving bonded
> > parameters to and from a metal center with the Hessian from a DFT
> > calculation. That may get you reasonable parameters for a totally bonded
> > model, and a fully bonded model would allow you to carry significant LJ
> > parameters on the Pt without disrupting the surrounding ligands (the LJ
> > interactions between Pt and bonded ligand atoms would be omitted by 1:2
> > exclustions).
> >
> > When you assign LJ parameters to metal ions, don't fall into the trap
> that
> > some have before: using a very small Lennard-Jones epsilon. What this
> does
> > is push back the point at which any significant repulsive force is felt,
> > until WHOOSH the r12 potential kicks in big time. It creates a "hard
> wall"
> > sort of effect that will hurt your energy conservation with standard MD
> > integrators. Rather like the computational equivalent of shaking a
> ketchup
> > bottle until you finally exceed the shear needed to thin the fluid and
> dump
> > it all over your plate. Use epsilons in the range 0.1 to 0.25 kcal/mol
> if
> > at all feasible.
> >
> > Dave
> >
> >
> > On Wed, Jun 29, 2016 at 4:34 AM, Zhenyu Meng <fdmm1989.gmail.com> wrote:
> >
> >> Dear David,
> >> Thanks for your information.
> >> I'm now working with a Pt complex interacting with DNA, I planed to
> >> constrain the distance between Pt and ligand atoms because as you said,
> no
> >> existing ff parameters for Pt.
> >> I also refer to UFF(universal force field) and in it there are LJ
> >> parameters for Pt, I wonder is that OK to use only the LJ parameters in
> >> UFF
> >> for Pt and GAFF for the rest of my complex?
> >>
> >> 2016-06-29 12:44 GMT+08:00 David Cerutti <dscerutti.gmail.com>:
> >>
> >> > The parameters there are the Lennard-Jones sigma divided by 2^0.83333,
> >> in
> >> > other words times sigma times 2^(1/6) and divided by 2, followed by
> the
> >> > Lennard-Jones epsilon. These are natural quantities to consider, as
> >> twice
> >> > the first number will give the location of the minimum, and the second
> >> > number will give the depth of the minimum. It's also numerically more
> >> > convenient to work with those two quantities rather than the raw sigma
> >> and
> >> > epsilon if you write out the formula and factorize it (for similar
> >> reasons,
> >> > the partial charges in Amber topology files are scaled by 18.2223, the
> >> > square root of the famous conversion factor of 332.0535).
> >> >
> >> > It looks like platinum in your frcmod, and yes, I can see how someone
> >> would
> >> > write in the comment "ATTN, needs revision." I don't think anyone has
> >> > decent Lennard-Jones parameters for platinum, and I doubt a
> >> general-purpose
> >> > pair of numbers truly exists.
> >> >
> >> > Dave
> >> >
> >> >
> >> > On Wed, Jun 29, 2016 at 12:26 AM, Zhenyu Meng <fdmm1989.gmail.com>
> >> wrote:
> >> >
> >> > > Dear Amber users,
> >> > > I wonder what's the meaning of values in NONBON section of frcmod,
> >> e.g.
> >> > > NONBON
> >> > > Pt 0.0000 0.0000 ATTN, need revision
> >> > > I check the website and the two values may be type RE(van der Waals
> >> > radius
> >> > > and the potential well depth parameters) or AC(The 6-12 potential
> >> > > coefficients), so how could I know which type these two values stand
> >> for?
> >> > > Your kind help will be highly appreciated!
> >> > >
> >> > > --
> >> > > Sincerely,
> >> > > Mr. Meng Zhenyu
> >> > > Division of Chemistry and Biological Chemistry
> >> > > School of Physical and Mathematical Sciences
> >> > > Nanyang Technological University
> >> > > _______________________________________________
> >> > > 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
> >> >
> >>
> >>
> >>
> >> --
> >> Sincerely,
> >> Mr. Meng Zhenyu
> >> Division of Chemistry and Biological Chemistry
> >> School of Physical and Mathematical Sciences
> >> Nanyang Technological University
> >> _______________________________________________
> >> 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
>
--
Sincerely,
Mr. Meng Zhenyu
Division of Chemistry and Biological Chemistry
School of Physical and Mathematical Sciences
Nanyang Technological University
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Wed Jun 29 2016 - 03:00:03 PDT
