Re: [AMBER] MMPBSA Electrostatic energy

From: Jason Swails <jason.swails.gmail.com>
Date: Wed, 11 Nov 2015 13:24:53 -0500

On Wed, Nov 11, 2015 at 11:58 AM, Asma Abro 28-FBAS/MSBI/F09 <
asma.msbi28.iiu.edu.pk> wrote:

> I am sorry for not making it clear.
> I meant that if electrostatic energy value is zero that the results I am
> getting are not valid/correct?
>

​It looks like your ligand is completely uncharged. This is what your
output file looks like:

Ligand:
Energy Component Average Std. Dev. Std. Err. of
Mean
-------------------------------------------------------------------------------
VDWAALS -5.9156 0.6699
 0.0424
ESURF 4.0138 0.0335
 0.0021

G gas -5.9156 0.6699
 0.0424
G solv 4.0138 0.0335
 0.0021

TOTAL -1.9018 0.6620
 0.0419

See how there is no EEL -- that's because it's 0 for every frame.



> And if the results are with zero value of EEL, then is there a problem?
>

​If your ligand has no charges, then clearly 0 is the correct result. Does
it make sense for all charges on all ligand atoms to be 0? For some
ligands it does, but I don't know anything about your ligand. If the
charges should *not* be zero, then you should go back and check your whole
simulation, as you will have to redo every step that was done with
incorrect charges.



> No, I am not using a solvent.
>

​MMPBSA or MMGBSA uses either PB or GB to model solvent implicitly.
MMPBSA.py is giving you a "solvation free energy", so there is clearly a
solvent somewhere. You should set the external dielectric constant equal
to the dielectric constant of whatever solvent you are trying to compute
the binding free energy in. 99.9% of the time, that's water, so you should
just leave the dielectric constants alone. If you want to calculate the
binding energy in something else, like cyclohexane, then it would be
appropriate to change the external dielectric (but in that case, only PB
really applies as GB was parametrized for water).

HTH,
Jason

-- 
Jason M. Swails
BioMaPS,
Rutgers University
Postdoctoral Researcher
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Wed Nov 11 2015 - 10:30:03 PST
Custom Search