#567 trouble with guanidinium

[david goodsell]

I've been using version 2.2.3 (Dec 1 09) on Linux to add hydrogens to a series of small molecules. It is working beautifully except for guanidinium and amidinium groups. In most cases, it treats these as uncharged, and I can't get the pH flag to make changes. Here is an odd example that shows a change in behavior:
PDB entry 1dwc--the ligand from the file is called MIT.
If I extract the ligand (attached here) and add hydrogens, the guanidinium comes out fine, and the pH function seems to work (the carboxyl gets protonated at low pH, and the guanidium becomes uncharged at high pH).
Then, delete the two oxygens in the carboxyl group (O33 and O34). With that modified file, the guanidinium is always treated as uncharged, at all pH values that I tried. I find this same behavior with almost all other files with guanidinium groups (for instance, ligand from pdb files: 1bmm, 1bmn, 1bra, 1dwb, etc).
Thanks for your help!
David Goodsell
goodsell@scripps.edu

[david goodsell]

[Geoff Hutchison]

Tim, could you take a look at this for the 2.3 release?