Hello,
I'll try to outline some differences below:
On Fri, Dec 16, 2011 at 5:21 PM, Qinghua Liao <fantasticqhl.gmail.com>wrote:
> Dear Amber users,
>
> Here are the results of the binding energy calculated by python and perl
> mm/pbsa respectively, and also the input files are showed here.
>
> the input file for calculation of perl version:
>
> .GENERAL
> VERBOSE 0
> PARALLEL 8
> #
> PREFIX snapshot
> PATH ../snapshots/
> START 1
> STOP 400
> OFFSET 1
> #
> COMPLEX 1
> RECEPTOR 1
> LIGAND 1
> #
> COMPT ../com.prmtop
> RECPT ../rec.prmtop
> LIGPT ../lig.prmtop
> #
> GC 0
> AS 0
> DC 0
> #
> MM 1
> GB 1
> PB 1
> MS 1
>
Note that MMPBSA.py does not have molsurf (MS 1), so this will cause some
differences.
> #
> NM 0
> #
>
> ################################################################################
> .PB
> #
> PROC 2
> REFE 0
> INDI 1.0
> EXDI 80.0
> SCALE 2
> LINIT 1000
> PRBRAD 1.4
> ISTRNG 0.10
> RADIOPT 1
> INP 1
> #
> SURFTEN 0.00542
> SURFOFF -1.008
> #
> IVCAP 0
> CUTCAP -1.0
> XCAP 0.0
> YCAP 0.0
> ZCAP 0.0
> #
>
> ################################################################################
> .MM
> #
> DIELC 1.0
> #
>
> ################################################################################
> .GB
> #
> IGB 5
> GBSA 2
> SALTCON 0.10
> EXTDIEL 80.0
>
MMPBSA.py does not give direct access to EXTDIEL. The default in MMPBSA.py
is 78.3, not 80 (which is the default value set in sander). We use the
default and don't allow it to be changed too easily because the GB models
here were parametrized to match PB results with those specific values of
extdiel and intdiel.
Also, some of the PB defaults are slightly different, I think (for
instance, I think fillratio is different in the 2 programs for the
MMPBSA.py input you gave).
If you really want to compare them with the same input parameters, use the
"use_sander=1" variable in MMPBSA.py and compare the _MMPBSA_gb.mdin file
with the corresponding sander input file generated by mm_pbsa.pl.
MMPBSA.py will actually give you the option of using a customized
_MMPBSA_gb.mdin (or _MMPBSA_pb.mdin) via the -use-mdins flag. So you can
tweak the MMPBSA.py input files to match the mm_pbsa.pl input files, which
should give you an answer that matches to the precision used by the
coordinate files.
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 Sat Dec 17 2011 - 16:00:02 PST
