From: Brian Radak (brian.radak_at_gmail.com)
Date: Tue Mar 26 2019 - 08:28:21 CDT

Psfgen does not yet support LONEPAIR statements inside topology/stream
files and thus cannot generate PSFs that contain lonepairs - we hope to
have this remedied in the next release. In the meantime, CHARMM-GUI is
probably the recommended workflow.

NAMD itself supports PSFs that contain most lonepair types (all of those
types which are currently utilized in the Drude force field and CGenFF).
However, it recently became known that certain (arguably undocumented)
changes in the RTF standard may require the addition of bonds (possibly
manually) between lonepairs and their parent atoms. The symptom for this
latter situation is a clear message stating "FOUND ISOLATED LONEPAIR".

HTH,
BKR

On Mon, Mar 25, 2019 at 11:01 PM Aravinda Munasinghe <aravinda1879_at_gmail.com>
wrote:

> Dear VMD users,
> I want to generate a molecule with a CL bond. And to get the topology
> information, I have used the new version of the parmchk (which uses CGenFF
> 4.0). However, since the new version supports the lonepair for halogens, my
> STR file has the following line
> LONEPAIR COLI LP1 CL C2 DIST 1.6400 SCAL 0.0
> However, when the psfgen reading the STR file, it throws the following
> error,
> psfgen) ERROR! FAILED TO RECOGNIZE LONEPAIR. Line 69: LONEPAIR COLI LP1
> CL C2 DIST 1.6400 SCAL 0.0
> Does this mean VMD psfgen still does not support lonepair implementation
> for NAMD input generation?. If that is so, what other ways can I
> incorporate the lone pair related topology to the psf file?.
>
> Best,
>
> --
> Aravinda Munasinghe,
>
>