[Bioperl-l] More on PDB and chains...

[Bernd Web]

Hi all,

The story about the chains is somewhat more complex.
HETATM can be part of the chain and present before TER. No problem.
We cannot just push all HETATMS after the TER to "non-chain". Some
records associate HETATM after TER with a chain. A structure like
chain (A,B,C etc), hetatm_chain (A,B,C etc) and non_chain (all other
HETATM) would be able to capture this.
An example PDB record where HETATMs after TER are labelled with a chain is 1QA7.

regards,
bernd

On 9/15/06, Bernd Web <bernd.web at gmail.com> wrote:
> Hi Jurgen,
>
> Thanks for your info.
> Re the chain labelling: I fully agree with you, in case the chains are
> labelled. HETATMS in the chain should remain in the chain. That's why
> we talked about non-chain hetero atoms.
> If the chain is not labelled which is the case if  only one (default)
> chain exists (e.g. 102L) all non-chain HETATMS also end up in the
> default chain. This is not nice: it becomes tedious to check if a
> HETATM is part of the chain or not. In case of such records (102L) it
> could be good not to store non-chain in the default chain and obey the
> TER chain termination label.
>
>
> Regards,
> Bernd