From: Gert Peters (
Date: Sun Nov 18 2012 - 12:50:32 CST

Dear all,

When trying to load a predicted RNA model vmd does not recognize this
model (pdb) as being a single nucleic acid strand. A dozen warnings
appear saying "Unusual bond between residues: ... (none) and ...

The pdb file is obtained from a RNA prediction algorithm and is poorly
defined when i comes to bonds. To improve the definitions in the pdb
file I parsed the pdb file through openbabel which removed most of the
warnings and improved the outcome but still there are some bonds being
unrecognized by vmd.

Now, because of these missing bonds the problem is that the nucleic
backbone I want to select isnt complete. Is there any way to enlarge the
distance searched in the distance search of vmd? Or is there another
solution to this problem? The only thing I want to do in vmd, at this
point, is compare rmsd of some structures (alot so manual editing of the
bonds in the pdb is impossible).

The most recent output for example:

Info) Analyzing structure ...
Info) Atoms: 1167
Info) Bonds: 1352
Info) Angles: 0 Dihedrals: 0 Impropers: 0 Cross-terms: 0
Info) Bondtypes: 0 Angletypes: 0 Dihedraltypes: 0 Impropertypes: 0
Info) Residues: 40
Info) Waters: 0
Warning) Unusual bond between residues: 8 (none) and 9 (nucleic)
Warning) Unusual bond between residues: 10 (none) and 11 (nucleic)
Warning) Unusual bond between residues: 21 (none) and 22 (nucleic)
Info) Segments: 1
Info) Fragments: 2 Protein: 0 Nucleic: 2

Thanks in advance,