Re: Getting extra atoms by Patching!

From: Kenno Vanommeslaeghe (kvanomme_at_rx.umaryland.edu)
Date: Thu May 08 2014 - 12:38:10 CDT

Please reply to the list rather than personal e-mail (hint to the list
admin: Majordomo has an option to override the Reply-To: address!)

On 05/08/2014 12:58 PM, Sadegh Faramarzi Ganjabad wrote:
> I replaced them with charges I got from semi-empirical
> calculations for the same molecule

Do NOT do that!!! Use standard charges.
http://mackerell.umaryland.edu/~kenno/cgenff/faq.html#esp
I already reworded that FAQ entry a dozen times, but every time, someone
comes up with a reason to think it doesn't apply to them.

> I'm wondering, if I want to use standard CHARMM amide atom types and
> charges, I also have to change names of those atoms in amino acid residues
> that are going to link together by patch.

The second column is the *name* and the third one is the *type*. They are
independent. In your patch, the *type* should be changed to match the
CHARMM amide types but the *name* should be the same as in the *PDB*.

Additionally, if the *name* in the patch is the same as the *name* in the
*residue before patching*, the type and charge of the existing atom are
simply modified. If it is different, a new atom gets created, which means
the patch should contain a corresponding DELETE command do remove the old
one. In fact, something like that might be your problem. Another possible
issue is that the "extra" atoms are simply hydrogens that are not resolved
in the PDB and that are placed at the origin by lack of the necessary
information to place them at the right position. CHARMM gets that
information from the IC table but I'm not sure what psfgen or VMD do.
There's an odd chance the problem might go away with fixing the atom types.

This archive was generated by hypermail 2.1.6 : Wed Dec 31 2014 - 23:22:23 CST