On Wed, Dec 19, 2012 at 11:20 AM, Jonathan Gough <jonathan.d.gough.gmail.com
> wrote:
> I ran a MMPBSA computation and the data isn't exactly what I expected. I
> am wondering if and what I am missing. Do these results make sense?
>
> protein - peptide MD was done in explicit solvent, 1 Na atom was used for
> charge balance.
>
> the total G is negative for the GB, but the PB is positive.
>
>
> |Input file:
> |--------------------------------------------------------------
> |Input file for running PB and GB
> |&general
> | interval=10,
> | verbose=1,
> |# entropy=1,
> |/
> |&gb
> | igb=2, saltcon=0.10
> |/
> |&pb
> | istrng=0.000, inp=1, radiopt=0
>
I would suggest making sure that the parameters between GB and PB are as
similar as possible when trying to make comparisons. For example, if you
set saltcon=0.1 in GB, set istrng=0.1 in PB.
Your non-polar solvation term looks worrying in PB. Since you're using
effectively the same method to treat nonpolar solvation in both PB and GB,
I would expect that term to be similar for both methods. In fact, they
should be identical as long as the SA method is the same (e.g., LCPO) and
the atomic radii are the same.
The ENPOLAR numbers don't make sense in your PB calculation (they are the
wrong sign, for one). I'm not sure what's happening there. I would
suggest substituting the ESURF values from GB into the ENPOLAR term for PB,
which would give you better agreement (since it's the ENPOLAR term that is
dominating your PB results).
Can you send me a couple _MMPBSA_*_pb.mdout.0 files to see what's happening
with that?
Thanks,
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 Wed Dec 19 2012 - 10:00:02 PST
