The gaps are large enough that I don't want to naively guess the
coordinates but do not want artificial charge - that's it.
The most easily automated solution seems to be:
1) let tleap guess the terminal coordinates
2) use parmed to delete the bond across the gap (ie add a TER character)
3) use parmed to reassign terminal atoms to ACE/NME atoms
4) write a new PDB and send it back through tleap
On Mon, Oct 18, 2021, 10:32 AM Jason Swails <jason.swails.gmail.com> wrote:
> Have you ever tried pdbfixer? It'll build-in missing loops via simple
> geometrical arcs.
>
> If you don't care about the physicality at the gaps (outside of not wanting
> terminal charges), you could always let tleap build the ridiculous bond
> then use ParmEd's `deleteBond` to remove that stretched bond (and all other
> valence terms that run through it). You'd probably want some kind of
> restraint to prevent the caps from going all over the place, but such an
> approach is likely easy enough to automate.
>
> All the best,
> Jason
>
> On Fri, Oct 15, 2021 at 1:53 PM Brian Radak <brian.radak.gmail.com> wrote:
>
> > I presume you are only recommending dangling ends for density fitting,
> > not simulation. Unfortunately I'm looking to prepare a ligand-protein
> > complex for a binding free energy calculation, so dangling ends are no
> > good, as are strong charged interactions with termini.
> >
> > I did notice the gap detection in pdb4amber, although most deposited
> > PDBs have a list of missing residues already, so this is not a huge time
> > saver. I played around with ParmEd for a bit trying to insert new
> > atoms/residues for the caps, but it was hopelessly difficult when the
> > insertion is in the middle of a chain (you have to re-index the residues
> > or else split the chain and then re-combine - just not good).
> >
> > I'm just surprised, bc this is ridiculously easy in CHARMM/XPLOR with
> > residue patching, even if the resulting capped residue is non-standard.
> >
> > Cheers,
> >
> > Brian
> >
> >
> > On 10/14/21 9:19 PM, David A Case wrote:
> > > On Thu, Oct 14, 2021, Brian Radak wrote:
> > >
> > >> Is there a script in AmberTools that automates insertion of capping
> > >> groups
> > >> (e.g., ACE and NME) along with TER characters when prepping a PDB for
> > >> tleap, etc.? I was kind of shocked when the manual suggested doing
> > >> this by
> > >> hand as the obvious way. Would this be an attractive feature in
> > >> pdb4amber
> > >> or are there simply enough third party preparation tools available on
> an
> > >> academic license?
> > >
> > > pdb4amber will tell you where the gaps are, but does not impose any
> > > particular
> > > way to handle them. I've never used ACE/NME in this situation, but
> > > that is
> > > just me. What phenix (and I) do with gaps is to use pdb4amber to find
> > > the
> > > gaps, then just add a deleteBond command to the tleap script to remove
> > > the
> > > bond that tleap creates during the loadPdb command. This leaves
> dangling
> > > ends, but that is not an obvious problem for molecular mechanics force
> > > fields. No need for TER cards in any pdb file, although they will
> > > appear if
> > > you later use ambpdb with the final prmtop file as input.
> > >
> > > Maybe others on the list have a more sophisiticated workflow here.
> > >
> > > ....dac
> > >
> > >
> > > _______________________________________________
> > > 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
> _______________________________________________
> 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 Mon Oct 18 2021 - 09:30:02 PDT
