From: Francesco Pietra (
Date: Fri Nov 14 2014 - 11:07:45 CST

I am sorry for having posted before looking at pdb better. There are many
atoms missing (entire side chains) from the experimental pdb. psfgen
reconstructed such chains.


On Fri, Nov 14, 2014 at 11:00 AM, Francesco Pietra <>

> With psfgen through vmd in text mode (charmm36 or 27), I started with all
> H-atoms present (neutron diffraction data). For a large number of H- and
> heavy atoms, psf gen "failed to set coordinates". Actually, in the pdb
> generated the coordinates are present, and the correct side chains are
> shown graphically for all atoms. However for those above the atom symbol
> (last pdb column) is not present.
> I wonder whether it is safe to go on with generated psf/pdb files without
> atom symbols or it would be better to remove all H atoms from starting
> files and let psfgen setting them. I don't see problems for the protein,
> however I also have organic ligands that required ad hoc parameterization.
> At any event, the location of H-atoms from accurate experiments is
> something valuable.
> Thanks for advice or indication where the point is clarified.
> francesco pietra