Hi all,
When I perform Poisson-Boltzmann calculations for MM-PBSA, I've noticed that
the value of ionic strength has virtually no effect on the calculated
electrostatic contribution. The maximum change I've observed is ~2 kcal/mol
between 0 and 150 mM. This is inconsistent with my intuition for the system,
as well as Generalized Born calculations, which change by 10's of kcal/mol.
I've tried these calculations with both the mm_pbsa.pl and MMPBSA.py scripts
to no avail. I'm using Amber 10 with all bug fixes applied (and all Python
script bug fixes applied). Here's a sample input for the Python script:
MMPBSA.py:
&general
startframe=2000, endframe=10010, interval=120, keep_files=2, verbose=2,
/
&gb
igb=5, saltcon=0.150,
/
&pb
istrng=0.150
/
Analyzing the verbose outputs, I cannot see any overt errors or warnings,
and the only difference I observe versus the zero ionic strength case is
that istrng/saltcon in the PB/GB .in files are changed to the appropriate
values (150 and 0.150, respectively).
On changing ionic strength, the only individual contribution that changes is
the E(GB) or E(PB), which of course is expected. Here's some sample output
of delta E(GB) and delta E(PB) for 0 mM and 150 mM:
IS 0 mM:
E(GB): 8197
E(PB): 8208
IS 150 mM:
E(GB): 8275
E(PB): 8206
As you can see, the GB energy gets less favorable (expected) but the PB
energy stays the same. I've tried this with a variety of salt concentrations
with no difference in behavior. I first observed this issue in some somewhat
exotic systems, but I've had the same problem even when calculating the
energy of melting of DNA, i.e. this is force field independent. (The DNA
melting calculations were for testing purposes only; I know they aren't
reliable indicators of, say, relative melting temperatures.)
I'm at a loss as to what I should do. Is it possible that I compiled Amber
without some important flag or additional program? Please let me know what
further details I can provide or where you think I should look.
Best,
Robert Elder
PhD Candidate
University of Colorado - Boulder
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Mon Feb 07 2011 - 13:30:10 PST
