thanks, that's a very helpful feature. of course people should think about
whether the remaining terms are still reasonable since it's a different
molecule now. for example, should the atom types or partial charges change?
I can imagine lots of cases where deleting a bond would be useful, but
other cases where people should probably generate a new prmtop from
scratch, depending on exactly what they are trying to learn. just adding
that here in case people read the archives...
On Thu, Sep 17, 2015 at 11:16 AM, Jason Swails <jason.swails.gmail.com>
wrote:
> On Thu, Sep 17, 2015 at 9:48 AM, Carlos Simmerling <
> carlos.simmerling.gmail.com> wrote:
>
> > Jason - correct me if I'm wrong, but if you delete the bond parameter,
> it's
> > true that it won't be there, but it won't delete the angle term between
> > atoms across the previous bond, or the torsion parameters. There is also
> > the problem of the nonbonded interactions. the atoms previously bonded
> > won't have any nonbonded, nor will the 1-3 atoms, so these molecules can
> > float on top of each other near the deleted bond. I would imagine the
> only
> > good way to do this is to remove the bond before you make the prmtop, not
> > after... unless parmed is smarter than I thought!
> >
>
> ​ParmEd is smarter than you thought ;). deleteBond does all of this. It
> deletes the bond in its entirety (including all other valence terms that
> include that bond), and it always recomputes the exclusions before writing
> a topology file.
>
> All the best,
> Jason
>
> --
> Jason M. Swails
> BioMaPS,
> Rutgers University
> Postdoctoral Researcher
> _______________________________________________
> 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
Received on Thu Sep 17 2015 - 09:00:03 PDT
