On Tue, Nov 3, 2015 at 1:11 PM, Gerald Monard <
Gerald.Monard.univ-lorraine.fr> wrote:
>
>
> On 11/03/2015 06:11 PM, Jason Swails wrote:
> > Moving this over to the developer's list
> >
> > On Tue, Nov 3, 2015 at 10:27 AM, Gerald Monard <
> > Gerald.Monard.univ-lorraine.fr> wrote:
> >
> >>
> >> OK. I got it now.
> >> So, now, instead of 'restraint = coordinate(...)', I can do:
> >> # d2-(d3-d4)
> >> &rst iat = 11,48, 48, 49, 49, 16,
> >> rstwt = 1,-1,1, r0 = 3.2, k0 = 0.0, /
> >> and it works just fine.
> >>
> >
> > Gerald -- am I correct in reading that this input works while the
> previous
> > input with the "natural language" input did not?
>
> Yes.
>
> >
> > If so, that raises a question as to what we should do here. Does anybody
> > use this input option, and is anybody familiar with the code that can
> look
> > into the issues and try to fix the bugs? If not, should we continue to
> > keep natural language input around when fixes to the problem involve
> > abandoning it altogether? (Does anybody use it on a regular basis?)
> Also,
> > pmemd does not support this input (and likely never will), so these input
> > files are sander-only.
> >
> > Thoughts?
>
> Several things:
> - I have not tested many things, but the problem I raised concerns the
> use of the 'coordinate' keyword (in "restraint = coordinate(...)"). It
> may/can/does work perfectly fine with other keywords.
> - I can't find any examples of "restraint = coordinate(...)" in the
> test/ and AmberTools/test/ directories. Moreover, the number of tests
> using the restraint keyword is very limited (2-3 tests at most).
> - The workaround that you suggested and that I now use (i.e., using
> rstwt) works fine as long as one wants to restrain distances. If one
> wants to restrain angles, planes, etc, that could be a potential problem.
>
There is nothing you can do with the "natural language" restraints that
you can't do with the iat, igr1-8, and rstwt variables to my knowledge.
Someone please correct me if I'm mistaken!
Torsions and angles are handled just fine.
All the best,
Jason
--
Jason M. Swails
BioMaPS,
Rutgers University
Postdoctoral Researcher
_______________________________________________
AMBER-Developers mailing list
AMBER-Developers.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber-developers
Received on Tue Nov 03 2015 - 10:30:05 PST
