VMD-L Mailing List

From: Andrew Dalke (adalke_at_mindspring.com)
Date: Tue Jul 23 2002 - 23:35:06 CDT

Hey all,

John Stone:
> VMD doesn't yet know how to process PDB conect records itself,
> but Andrew Dalke wrote a script that will create a PSF from a PDB with
> conect records, its in the TB web site in the "MDTools" area if I recall
> correctly),

That program is called 'conect2psf'. I occasionally get requests for
this program, including one which came in last week, so let me be
definitive about it.

I wrote that, as I recall, during Spring Break '96 (or maybe '95). It
was written in a C++ toolkit I never released, and I don't even know
where that toolkit is now. Not only that, but I only ever released a
binary for IRIX 5.x.

As I recall, the format of the PSF file was a bit tricky to reverse
engineer. I don't remember it any more but I bet someone at the TB
group knows at least enough to rewrite that code in Python/ Tcl/ Perl.
I guarantee it will be much easier to understand then whatever I did
nearly a decade ago.

In addition, I once wrote a little hack in VMD's Tcl to modify the
topology given the CONECT records in a PDB after the file was loaded.
This was in a VMDTech which is available from the vmd-l archives, at

  http://www.ks.uiuc.edu/Research/vmd/mailing_list/vmd-l/0862.html

It contains plenty of detail on how that approach works and on
the limitations of that hack.

  That's about all I can suggest for now. (Unless you want to hire
my services to write that code :) Hopefully others on the list can
provide some better assistance.

                                        Andrew
                                        dalke_at_dalkescientific.com