It's really difficult to read other people's source code when
converting something from one language to another especially porting to an object oriented language.
It would be much easier though to just start from the correct equation --with the adjustment conditions if I can determine what they are.
Is the correct equation out there some where. I'm primarily interested in nucleic acids with (I'm learning) implied solvation?
Should I start from the wikipedia entry and make all the adjustments and modfications to the code you mentioned.
I don't think I'm going to be that successful digging through the source code unless it's superbly documented.
What is your recommendation?
Thanks for taking the time to enlighten me of all the problems and pitfalls.
Dan
-----Original Message-----
From: Ross Walker <ross.rosswalker.co.uk>
To: 'AMBER Mailing List' <amber.ambermd.org>
Sent: Mon, Mar 15, 2010 7:16 pm
Subject: RE: [AMBER] Calculating Amber Force fields from Wikipedia
Note, the Wikipedia entry, like most published examples of the AMBER force
field, including the original Cornell et al JACS paper, is incorrect and
should not be relied upon for implementing the AMBER force field. You would
be far better referring to the source code from either SANDER, NAB or
something like Tinker.
The entry you refer to here is for a gas phase simulation. For an implicit
solvent or explicit solvent calculation the form of this equation is
substantially more complicated in terms of the non-bonding interactions.
Additionally it states:
"Fourth term (double summation over i and j): represents the non-bonded
energy between all atom pairs, which can be decomposed into van der Waals
(first term of summation) and electrostatic (second term of summation)
energies."
double summation of I and J. This text is wrong. One does not double count
but sums from i=1,j; i<j. If one did double count it would be necessary to
have a factor of one half in front of the double sum.
There is also no mention here of the fact that nonbond interactions are not
counted in the case where I is bonded to J, or I forms an angle with J and
that where I and J form atoms 1 and 4 of a dihedral the electrostatics are
divided by 1.2 (for FF99) and the VDW by 2.0 (for FF99).
Thus the nonbond part of the equation should have the sum with the range
i-1,j; i<j where i,j are not part of the set of bonds, angles and dihedrals.
And then have an additional summation over unique dihedrals (1-4's) that
provides the 1.2 and 2.0 scaling.
Note, just about every text book and paper discussing the AMBER force field
equation is incorrect in this regard hence why the source code is really
your definitive definition here. Unfortunate but true.
Good luck,
Ross
> -----Original Message-----
> From: amber-bounces.ambermd.org [mailto:amber-bounces.ambermd.org] On
> Behalf Of Whimsica.aol.com
> Sent: Monday, March 15, 2010 8:33 AM
> To: AMBER Mailing List
> Subject: Re: [AMBER] Calculating Amber Force fields from Wikipedia
>
> The equation on wikipedia has electostatics torsions angles bonds...
> Parm99 has only 2-3 parametrs and they are not labeled. Same with the
> allnuc94 how do I tell which number corresponds to which parameter in
> the wikipedia equation
>
> Thanks for any guidance
>
> Dan
>
> On Mar 15, 2010, at 7:47 AM, Carlos Simmerling
> <carlos.simmerling.gmail.com
> > wrote:
>
> > they are available for download on the amber web site
> > (www.ambermd.org),
> > along with a lot of information on the format. make sure to compare
> > energies
> > and forces from your code to that from Amber- there are multiple
> > examples in
> > the past of people using "Amber" force fields, but not doing the
> > conversion
> > or calculation correctly.
> >
> > On Mon, Mar 15, 2010 at 8:46 AM, <whimsica.aol.com> wrote:
> >
> >>
> >> I'm writing a small molecular dynamics system for students and
> >> would like
> >> to try using amber forces in a simulation.
> >> On wikipedia it says that the amber force field can be calculated
> >> using
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >> and the parameters of the force field (e.g. force constants,
> >> equilibrium
> >> bondlengths and angles, charges). Peptide, protein and nucleic acid
> >> parameters are provided byparameter sets with names beginning with
> >> "ff" and
> >> containing a twodigit year number, for instance "ff99"
> >>
> >> Could someone please point me to the parameter file(s) that would
> >> including
> >> everything necessary to use this equation.
> >> I'm writing a small molecular dynamics system for students and
> >> would like
> >> to try using amber forces in a simulation.
> >>
> >> Thanks,
> >>
> >> Daniel Savage
> >> CinematicSciences
> >>
> >>
> >> _______________________________________________
> >> 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
_______________________________________________
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 Mon Mar 15 2010 - 20:00:03 PDT
