On Wed, Sep 21, 2011 at 1:56 PM, Marek Maly <marek.maly.ujep.cz> wrote:
> Dear all,
>
> I have question about SCEE parameter.
>
>
> I recently did MM/PBSA analyses of my system using mm_pbsa.pl script .
> I set variable VERBOSE to 1 to have possibilities to check all parameters
> in intermediate files.
>
> For calculation of entropy I used nmode program. What surprised me here
> was that in output files
> for sander minimization (sandmin*.out) generated during the calculation I
> found different value for
> scee parameter than in case of output files for normal mode calculation
> (nmode*.out).
>
> Below are sandmin*.in and nmode*.in files and relevant parts of both
> output files mentioned above.
>
> As you can see in sandmin*.out is scee=1.2 while in case of nmode*.out
> there is scee=2.0.
>
Sure enough... Line 51 of rdinp.f:
rdinp.f:51: scee = 2.0
For some reason the default value for scee is set to 2.0, yet the amber
force field was parametrized with scee=1.2. However, nmode is not an
actively maintained program, and there are alternatives to using it for
normal mode calculations (nab programs, for instance).
However, the issue of the scaling factors is complicated in amber. In
sander and pmemd, you're allowed a mixed-scaling option in which different
dihedrals can be treated with different scaling constants, primarily to
enhance the cooperativity of GLYCAM and Amber force fields. However, no
other programs can handle that currently (including nmode, nab, and I
believe mdgx as it's released with AmberTools 1.5). Therefore, if you get
the "correct" scaling treatment for your system in sander/pmemd, then you
will get the wrong (and different) scaling in nmode, since nmode doesn't
support mixed scaling (so everything will be set to 2.0 and 1.2, or 2.0
depending on what you set). Therefore, without jumping into the code a
little bit on your own, there's no way to get entropy contributions using
mixed-scaling.
I would, however, suggest against using nmode for entropy calculations.
MMPBSA.py can most likely do what you're trying to do, and it also uses the
more supported and more advanced nab NMODE routines for entropy
calculations.
> To be frank I would assume the same value of scee parameter during all
> stages of the normal mode calculation,
> more likely value 1.2 as in my system GAFF (scee=1.2), ff99bsc0
> (scee=1.2), Glycam_06 (scee=1) forcefields are used.
>
> So if nmode scee default value scee=2.0 ? should be changed in my case,
> how to do it ? I did not find any information
> about setting this parameter using standard mm_pbsa.pl input file. I guess
> it could be done by writing this parameter
> directly in nmode*.in file which is by default created by mm_pbsa.pl
> script. If I am right, how is then possible to correctly run
> mm_pbsa.pl with "by hand" created sandmin*.in and nmode*.in files instead
> of standard mm_pbsa input file ?
>
You could do this, or change the nmode source code above to default to 1.2
by adjusting the line I pointed out above and recompiling.
>
> I would like also to know why sander printed "Note: 1-4 EEL scale factors
> were NOT found in the topology file.".
>
> Is it because I am using several ffs which differ in scee (scnb)
> parameters and for my input files using tleap ( AmberTools 1.4 )
> instead of sleap ?
This was answered in my previous email.
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 Wed Sep 21 2011 - 12:00:05 PDT
