On Wed, Dec 7, 2011 at 5:15 AM, Marc van der Kamp
<marcvanderkamp.gmail.com>wrote:
> Hi Jason,
> Thanks again. I had diffed the files as well, and based on that thought
> energies should indeed be zero with exclusions2.prmtop.
> My first guess as to why your input gives the expected result, whereas ours
> didn't was right:
> When you remove igb=6 from the input (effectively setting igb=0), then the
> energies aren't zero and identical as with the original prmtop.
> I hadn't realized that for this small test system, igb=0 and igb=6 wouldn't
> be equivalent.
> BTW, I tried setting "igb=0, eedmeth=4", which should be the true
> equivalent of igb=6, but eedmeth wasn't understood:
>
eedmeth is a variable in the &ewald namelist. However, setting that
variable gives the same (wrong?) behavior you described above. I'm not
sure why igb=0,eedmeth=4 gives different results in this case than igb=6.
It appears that igb=0,eedmeth=4 does not obey the exclusion list, but
igb=6 does. (When I add bonds, angles, and a dihedral to the system,
igb=0,eedmeth=4 does properly exclude each atom from the non-bonded
calculation list).
Note that these 2 approaches follow different code paths. The first
(igb=0,eedmeth=4) traverses the ewald code for the EEL/VDW calculations,
whereas igb=6 traverses the GB code path for the EEL/VDW calculations, so
it would appear that only GB calculations will respect the
arbitrarily-modified exclusion list.
You can see this when igb=1 (or any other GB model), you'll get the result
you're aiming for (note, however, that the EGB energies don't change based
on the exclusion list).
HTH,
Jason
>
> Cannot match namelist object name eedmeth
> namelist read: missplaced = sign
>
> All the best,
> Marc
>
> On 7 December 2011 01:07, Jason Swails <jason.swails.gmail.com> wrote:
>
> > On Tue, Dec 6, 2011 at 3:35 PM, Marc van der Kamp
> > <marcvanderkamp.gmail.com>wrote:
> >
> > > Hi Jason,
> > >
> > > Yes, I think the student originally put them in one command, but played
> > > around as it didn't seem to work.
> > > I get the same prmtop (only timestamp differs) when I do use the
> combined
> > > command, and the correct atoms (all 4 ions) are selected by the mask.
> > >
> >
> > When I diffed the 2 topologies you gave me, I got this:
> >
> > $ diff -w 4ions.prmtop exclusions2.prmtop
> > 1c1
> > < %VERSION VERSION_STAMP = V0001.000 DATE = 12/06/11 15:38:03
> >
> > ---
> > > %VERSION VERSION_STAMP = V0001.000 DATE = 12/06/11 13:59:27
> > 8c8
> > < 4 4 0 0 0 0 0 0
> 2
> > 0
> > ---
> > > 13 4 0 0 0 0 0 0
> 2
> > 0
> > 25c25
> > < 1 1 1 1
> > ---
> > > 6 4 2 1
> > 91c91,92
> > < 0 0 0 0
> > ---
> > > 2 2 3 3 4 4 3 3
> 4
> > 4
> > > 4 4 0
> >
> > Clearly the exclusion atom list is different here. (I've fixed parmed so
> > it doesn't duplicate identical exclusions like above (2 2 3 3 4 4 3 3 4
> 4 4
> > 4 0 should be 2 3 4 3 4 4 0) but the results from sander remain the
> same).
> >
> > >From the original prmtop I get the following in my mdout file:
> >
> > NSTEP ENERGY RMS GMAX NAME NUMBER
> > 1 1.0587E+07 4.5553E+07 1.1102E+08 Cl- 2
> >
> > BOND = 0.0000 ANGLE = 0.0000 DIHED =
> > 0.0000
> > VDWAALS = 10587886.4492 EEL = -469.5927 EGB =
> > 0.0000
> > 1-4 VDW = 0.0000 1-4 EEL = 0.0000 RESTRAINT =
> > 0.0000
> >
> >
> > >From the exclusions2.prmtop I get:
> >
> > NSTEP ENERGY RMS GMAX NAME NUMBER
> > 1 0.0000E+00 0.0000E+00 0.0000E+00 Na+ 1
> >
> > BOND = 0.0000 ANGLE = 0.0000 DIHED =
> > 0.0000
> > VDWAALS = 0.0000 EEL = 0.0000 EGB =
> > 0.0000
> > 1-4 VDW = 0.0000 1-4 EEL = 0.0000 RESTRAINT =
> > 0.0000
> >
> >
> > So I'm not quite sure why you were getting what you were getting. Here's
> > my mdin file:
> >
> > input file
> > &cntrl
> > imin=1, maxcyc=1, ntb=0, igb=6, cut=9999,
> > /
> >
> > Thanks in advance for your help!
> > > Marc
> > > PS I note that exclusions2.prmtop is a lot smaller than the original
> > > 4ions.prmtop, but I don't think this is the problem
> > >
> >
> > This is due to whitespace. leap prints every line in the prmtop file
> with
> > 80 characters (filling the empty ones with spaces). parmed does not have
> > those spaces, which is why the file size is smaller.
> >
> > All the best,
> > 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
> >
> _______________________________________________
> AMBER mailing list
> AMBER.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber
>
--
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 07 2011 - 06:00:10 PST
