[phenixbb] Dummy atoms (was: useful PDFs by Pavel)

Frank von Delft frank.vondelft at sgc.ox.ac.uk
Wed Jul 28 04:31:21 PDT 2010


Hi Pavel

A bunch of extremely useful slides, thanks for making them available to 
the world!

Inevitably, though, given the nature of our field, there's one slide I 
must challenge, the 2nd-last one of the validation pack, where you 
recommend against the use of UNK atoms, but don't say why:

<snip>
Some programs and people tend to interpret unknown density using "dummy
atoms". In PDB files it typically looks like this:
ATOM 10 O UNK 2 6.348 -11.323 10.667 1.00 8.06 X
ATOM 11 O UNK 2 6.994 -12.600 10.740 1.00 7.16 X
ATOM 12 O UNK 2 6.028 -13.737 10.607 1.00 6.58 X
ATOM 13 DUM UNK 2 6.796 -15.043 10.583 1.00 8.28
ATOM 14 DUM UNK 2 5.099 -13.727 11.792 1.00 7.15
- *Do not deposit this in PDB*, especially if chemical element type is 
undefined
(rightmost column)
</snip>


Why should one not be allowed to indicate in the model that there was 
very clear, atomic density whose chemistry nevertheless could not be 
explained?

The reason we group atoms into chemical entities is purely so we can 
impose restraints on interatomic distances and thereby compensate for 
the poor data-parameter ratio.  But chemistry is not a substitute for 
making /scientific/ sense of the model, that problem frequently lies 
well beyond the reach of the model -- yet that should not stop the model 
from being deposited, as would be the logical conclusion of your 
recommendation.

The scenario is universal, particularly acute in structural genomics but 
possible even in ligand-binding studies.

(Of course, if what you meant to say that UNK atoms are to be used 
extremely judiciously, and it is *not* okay to flood a bad model with 
UNK atoms only to get the R-factor down -- then I'm totally with you!)

phx.

_Full disclosure_:  I am one of the culprits behind the JCSG persuading 
the PDB to accept UNK as valid residue type, after we had run into 
simply too many models where there was clearly something bound, but no 
quick/cheap way of figuring out what it was.




On 27/07/2010 20:43, Pavel Afonine wrote:
>  Hi Everyone,
>
> recently I had to update or do from scratch my talk "slides" on 
> various subjects such as crystallographic structure refinement, maps, 
> validation and some general PHENIX overview. Since I spent pretty 
> significant amount of time making all these "slides" , ... I was 
> thinking that may be this might be of use for some of you, so here it 
> is...
>
> - 42 pages of general introduction to structure refinement:
> http://www.phenix-online.org/presentations/latest/pavel_refinement_general.pdf 
>
>
> - 45 pages of phenix.refine overview (including extended details about 
> its use from the command line):
> http://www.phenix-online.org/presentations/latest/pavel_phenix_refine.pdf
>
> - 42 pages of "Some Facts About Maps":
> http://www.phenix-online.org/presentations/latest/pavel_maps.pdf
>
> - 50 pages of "Crystallographic Structure Validation":
> http://www.phenix-online.org/presentations/latest/pavel_validation.pdf
>
> - 31 pages of introduction to PHENIX:
> http://www.phenix-online.org/presentations/latest/pavel_phenix_intro.pdf
>
> Most of the slides in "Introduction to PHENIX" came from Paul Adams, 
> Tom Terwilliger and Nat Echols. Thanks Nat and Jeff Headd for 
> providing with interesting examples of structures with unusual geometry.
>
> All the best!
> Pavel.
>
> _______________________________________________
> phenixbb mailing list
> phenixbb at phenix-online.org
> http://phenix-online.org/mailman/listinfo/phenixbb
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://phenix-online.org/pipermail/phenixbb/attachments/20100728/1c2fc6e7/attachment-0003.htm>


More information about the phenixbb mailing list