Discussion List Archives

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: Ambiguity in atom_site.disorder_group value -1

Hi Bob,

 

 

I am thinking in terms of describing "local disorder groups" -- the pairing -- would be akin to the ability in a CIF file to describe bonding patterns, which (I think) are modeling aspects -- interpretation-based, not refinement-based. Does that sound right?

 

I would agree that chemical bonding, which CIF has very strong support for representing, is primarily a matter of interpretation of the structure model, not of refinement.  However, many refinement packages have the ability to use refinement restraints and constraints that are functions of (perceived) bonds, so bonding is not exclusively a matter of model interpretation.

 

The selection and population of disordered sites is a matter of refinement, but the details of such refinements frequently involve restraints derived from an interpretation of the disorder in terms of two or more specific molecular conformations or orientations.  So again, not a complete separation of refinement from interpretation.

 

When disorder components are crystallographically independent of each other, pairings are already pretty evident from standard CIF items that are typically reported.  Disorder group numbers, site occupancies, (computed) intersite distances, and (absence of) bond records all speak to this, but there is no item in core CIF that explicitly says that two or more groups of sites are _each others’_ alternatives.  Disorder group numbers come close, though.  When disorder components are crystallographically dependent on each other, negative disorder group numbers convey that information.  But as Brian’s example shows, that does not necessarily provide a complete picture.

 

 

it would be great if the depositor could indicate their interpretation as well and so show that as a default or by  some sort of selection.

 

I do not disagree that it would be great if depositors could and did represent those details in CIF.  But I hesitate to commission work on the “could” – which I expect would be a significant amount with a relatively complex result – when I am doubtful that the “did” would follow.  I anticipate that many crystallographers and crystallographic programmers would be of the opinion that what we already have in that area (disorder group numbers, etc.), though imperfect, is good enough.  But perhaps I would be surprised.

 

 

Best regards,

 

John

 

From: Robert Hanson <hansonr@stolaf.edu>
Sent: Monday, October 31, 2022 12:54 PM
To: Distribution list of the IUCr COMCIFS Core Dictionary Maintenance Group <coredmg@iucr.org>
Cc: Bollinger, John C <John.Bollinger@STJUDE.ORG>
Subject: Re: Ambiguity in atom_site.disorder_group value -1

 

Caution: External Sender. Do not open unless you know the content is safe.

 

Thank you, John. That was very interesting to me. So, perhaps, what I am thinking in terms of describing "local disorder groups" -- the pairing -- would be akin to the ability in a CIF file to describe bonding patterns, which (I think) are modeling aspects -- interpretation-based, not refinement-based. Does that sound right?

 

Jmol can certainly show any subgroup of these groupings, but it would be great if the depositor could indicate their interpretation as well and so show that as a default or by  some sort of selection.

 

Bob

 



Email Disclaimer: www.stjude.org/emaildisclaimer
Consultation Disclaimer: www.stjude.org/consultationdisclaimer
 
_______________________________________________
coreDMG mailing list
coreDMG@iucr.org
http://mailman.iucr.org/cgi-bin/mailman/listinfo/coredmg

[Send comment to list secretary]
[Reply to list (subscribers only)]