On Fri, Jun 28, 2013 at 2:11 AM, Asmita Gupta <asmita4des.gmail.com> wrote:
> Dear users,
>
> Are the energy terms given in amber .out file and those calculated by
> MMPBSA analysis equivalent?
>
Typically no. They are equivalent ONLY if you run your original simulation
in implicit solvent (the same model you use for your MM/PBSA analyses).
For RNA, you're almost exclusively restricted to running your calculations
in explicit solvent if you want something reasonable to come out.
As a result, the energies correspond to completely different simulations
(implicit solvent gives a solvation free energy with infinite dilution,
whereas explicit solvent calculations give a series of single-point
energies with high solute concentrations).
> As far as i understand, .out file energies are composed of potential and
> kinetic energy terms. We usually take a cut off of 10 Ang for all non
> bonded interactions. So does .out energies contains solvation energy and
> entropy component as well or is it just the MM energies?
>
The bonded energies will be the same unless your solvent has bond, angle
and dihedral terms that are not omitted due to the use of SHAKE or SETTLE,
but that's it. (They actually won't be *exactly* the same, because the
snapshots printed in the trajectory are a time-step ahead of the snapshots
for which the energies were computed in the mdout file). The vdW and EEL
energies will contain contributions between solvent atoms and
solvent-solute atoms which will obviously not appear in MM/PBSA
calculations.
Basically, the underlying potential is completely different.
My doubt arose because if one wants to do stability calculations in case of
> single RNA species and its single point mutants trajectories, what should
> be the correct approach?
>
This is a very challenging question since implicit solvent models tend to
do poorly with nucleic acids. You could do restrained MD and only allow
the water to move so you get an energy distribution in each conformation,
then compare the relative energies between the two systems (without the
restraint energy, of course). Or you could use MMPBSA.py with the PB or
RISM solvent models to do a stability calculation.
Furthermore, i had some Mg2+ ions in the original crystal structure, and 2
> of them directly contact RNA backbone. Should i treat these ions as ligands
> , in case of MMPBSA or it is better to remove them?
>
Do you want to get their binding free energy? Or do you want the relative
stability? If you want the binding energy, you treat them as ligands. If
you want the relative stability between two conformations, include them as
part of the receptor.
However, implicit solvent models are not known to treat ions particularly
well, since that's not their intended purpose (PB includes ionic strength
effects via the Boltzmann part of the Poisson-Boltzmann equation). If the
ion is structurally important, then you probably need to explicitly include
it. However, what you're trying to do is probably outside the accuracy
envelope of MM/PBSA calculations. I would encourage you to look at what
previous studies have done in this case (or be prepared to do some
extensive validation of whatever approach you choose).
HTH,
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 Fri Jun 28 2013 - 08:00:02 PDT
