On Tue, 25 Mar 2014 09:48:53 -0400
David A Case <case.biomaps.rutgers.edu> wrote:
> On Mon, Mar 24, 2014, Hannes Loeffler wrote:
> >
> > I just realised where the actual problems is. Leap does a valency
> > check by default and refuses to accept more than one bond per
> > hydrogen. It will keep the very first bond it encounters in the
> > input and discard others that may have been defined. I have
> > overlooked this and accidentally created the same bond twice (after
> > set pert true).
> >
> > I wonder though why leap would allow this when this will result in a
> > fatal error on parmtop writing anyway. Is there a valid use case
> > for this or shouldn't this better be captured at the time a user
> > tries to create the same bond again?
>
> Can you (re-)send a test case that shows the problem, based on your
> recent understanding of what is happening? I think the "pert" flag
> in LEaP is no longer relevant, and should never be set, and we might
> not be catching errors that involve its use. But it's not clear
> whether or not "pert" is involved in your problem or not.
>
> Thanks for sending these reports, and for finally finding the cause
> of the original problem.
See MORPH1.mol2 in the attachment. Leap keeps the first bond for H it
encounters, that is to DU6. This "problem" came up as part of my
FESetup script because I naively assumed that leap would keep the
"natural" bond to C. I accidentally created the bond DU6-H again. So
the following script would show DU6-H as the only bond.
source leaprc.ff12SB
source leaprc.gaff
mods = loadAmberParams MORPH1.frcmod
s = loadmol2 MORPH1.mol2
desc s.1.H
The work around is to delete every bond on suspicion because deletebond
on a non-existent bond will only result in a message on stdout and is
thus harmless.
set s.1.H pert true
deletebond s.1.H s.1.DU6
bond s.1.H s.1.DU6
deletebond s.1.H s.1.DU10
bond s.1.H s.1.DU10
deletebond s.1.H s.1.C
bond s.1.H s.1.C
I suppose when you say pert is no longer relevant you refer to the way
free energy calculations are now done with AMBER? I still use this
feature because it is convenient for support of other MD software. I
just have to find out now what the best way is to "copy" bonded
parameters between the states when they are otherwise zero because of
dummies. Direct manipulation of the parmtop through parmed seems
promising. To support free energy calculation in pmemd I guess I have
to do learn this anyway.
Many thanks,
Hannes.
--
Scanned by iCritical.
Content-Transfer-Encoding: base64
Content-Disposition: attachment; filename=MORPH1.mol2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_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Tue Mar 25 2014 - 08:00:02 PDT
