On Fri, Aug 12, 2011 at 6:27 PM, manikanthan bhavaraju <
manikanthanbhavaraju.gmail.com> wrote:
> Dac I am using ff99SB forcefield.
>
ff99SB is a protein force field primarily. It uses ff99 + (very good!)
backbone torsion modifications to fix ff99's problems. This force field was
developed in 2006. The nucleic acid part of this force field is still fresh
ff99 (and maybe no changes even since ff94?) In this case, the force field
that you're using is from 1999, which is over 12 years old at this point.
I would suggest using a newer force field (see ff10 in AmberTools 1.5)
unless you have a specific reason for using one that's so old (such as some
type of comparative study with old results). Reviewers would probably
question why you used such an antique force field for your simulation when
newer ones were so readily available (and it would call into question the
validity of your results).
> But I have a question regarding the DNA and DNA + ligand implicit
> solvations
> simulations.
>
Search the literature here. Implicit solvent is notoriously poor when
treating nucleic acids (nucleic acid simulations play *very* strongly to
GB's and even PB's weaknesses). Again, it's questionable how much you can
learn doing nucleic acid simulations in implicit solvent using such an old
force field.
>
> When we are preparing prmtop and inpcrd files in both cases, Do we have to
> add counterions ? In my case the DNA has -22 charge and ligand +1. Can
> counterions stablize the simulations?
>
As Dan Roe pointed out, neutralize if you plan on doing explicit solvent
calculations, but otherwise implicit solvent models take account of simple
counterions via the Boltzmann part of the Poisson-Boltzmann equation or via
a Debye screening parameter in the GB equation.
Nucleic acid simulations are notoriously difficult compared to protein
simulations, which is why there is such a smaller number of people working
on them (and the protein force fields are so much more thoroughly tested).
Furthermore, the barrier to starting a successful nucleic acid project is
steep, and you will probably have to do a large number of experiments to
find out what works for you (i.e. what water model do you use, TIP3P, TIP4P,
TIP4Pew, implicit solvent, etc.) and do a lot of reading to see what
approaches other scientists had the most success with.
Best of luck with this,
Jason
--
Jason M. Swails
Quantum Theory Project,
University of Florida
Ph.D. Candidate
352-392-4032
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Sat Aug 13 2011 - 12:30:02 PDT
