Discussion List Archives

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

(80) Multipole coefficients; mailing lists; coreDMG

  • To: COMCIFS@iucr
  • Subject: (80) Multipole coefficients; mailing lists; coreDMG
Dear Colleagues

This circular, the last for 1997, includes contributions from David
Brown (D>), Howard Flack (H>) and George Sheldrick (G>).

Existing discussion threads
===========================
D77.1 Multipole population coefficients
---------------------------------------
D> 	Mark's biases are showing here.  It may be that there are many
D> different conventions in use, but surely some are more often used than
D> others.  There is nothing to stop us from defining more than one
D> convention, but the act of defining these conventions should help to
D> persuade the community to start talking the same language.  We have had a
D> reasonable request to provide a definition for multipole items.  We could
D> do this on our own, or we could ask the Commission on Charge, Spin and
D> Momentum Densities to provide us with one or more definitions that would
D> answer in most situations.  It is their task to provide some kind of
D> standardisation in the field.  If they are unwilling to recommend
D> definitions, we may be forced to do the job ourselves, which I am sure
D> would be much less satisfactory.  Perhaps Mark can raise this matter with
D> the Commission. 

H> M> 1.   The issue I've always felt most strongly about is the appropriate
H> M> archival of DATA! To me, the most useful information for charge density
H> M> purposes is the experimental observations and the associated conditions
H> M> of the experiment.  
H> 
H>  I concur utterly and entirely.
H> 
H> D> The present dictionary takes care of all of that I think
H> 
H>  I disagree. 
H> 
H> M> (although it is clear to me that too few of the charge density
H> M> community actually deposit their data in CIF format anyway).
H> 
H>  It's not only the electron density people who are amiss. Look at the
H> "structure factor CIFs" displayed for Acta Cryst C to have a feel for
H> the diversity.
H> 
H>   BUT ITS THE RAW DATA AND CONDITIONS THAT IT IS ESSENTIAL TO HAVE
H> AVAILABLE. EVERYTHING ELSE IS SEEN THROUGH THE EYES OF A MODEL.
H> 
H> D> dipole vectors for each atom).  There also seems to be an opportunity
H> D> for the writers of structure refinement packages to allow raw
H> D> intensities to be read and written in cif so that different refinement
H> D> packages can be used. With few additions, cif would perform
H> D> this function well.  I am surprised that no one seems to have thought
H> D> of this application.
H> 
H>   Take a look at J. Appl. Cryst. (1992) 25, 455-459.

D77.2 Absolute structure
------------------------
I include some discussion from David Brown, and an enhanced definition from
Howard Flack. David has proposed that discussions of this type should be
moved in future to the mailing list that will soon be established for the
core Dictionary Maintenance Group (see D80.2 below), and I shall do that as
soon as I have time to discuss some of the details of managing that list
with David and to establish it properly - this will probably be in a couple
of weeks time. Meanwhile I post the continuing discussion here for the
record.

D> 	Howard and I seem to have quite different ideas about what is
D> expected in the _reflns_number_* fields.  I would have thought that
D> *_total would be a number in the thousands, counting every reflection that
D> was fed separately into the least squares program.  That is the way I read
D> the definition.  On the other hand, I find myself somewhat perplexed about
D> what is meant to go into *_Friedel.  I would assume that it would be the
D> number of reflections that would be left out of the refinement if the
D> center of symmetry was assumed, i.e. it would be roughly a half of
D> *_total.  If this is the case, I do not see the point of this item.  If
D> Howard is correct, then the number must always be either 1 or 2 and this
D> is probably not very interesting either.  There may be some virtue in
D> listing the number of sets of equivalent reflections that are measured, or
D> that are included in the refinement (the latter always being less than or
D> equal to the first) to indicate the extent of the redundancy in the use of
D> symmetry equivalent reflections, but in practice the different sets are
D> not likely to contain the same number of reflections.  Do we say that 2
D> sets were included in the refinement when one set contains a single
D> reflection?  I would suggest that questions relating to the definition of
D> crystal class and Laue symmetry be referred to symCIF and that whoever
D> proposed *_Friedel explain what they had in mind and why this is a useful
D> item to include.
D> 
D> 	I like Howard's definitions of _chemical_absolute_configuration
D> and _chemical_optical_rotation, and we should also consider renaming the
D> items _refine_ls_abs_structure_* to _refine_ls_absolute_configuration_*
D> with Howard's enhanced definition. I am becoming more and more convinced
D> that the names we use should be as precise as we can make them - anything
D> less is only likely to cause confusion, either because the item has not
D> been properly thought out, or someone will want to define a different item
D> that logically should have the name we have already used (incorrectly) for
D> something else. 

H>   After e-discussion with Tony Linden, I have added two extra
H> _enumeration_detail's for _chemical_absolute_configuration.

These new options are syn and unk. The complete definition is now:

H> data_chemical_absolute_configuration
H>     _name                      '_chemical_absolute_configuration'
H>     _category                    chemical
H>     _type                        char
H>     loop_ _enumeration           
H>           _enumeration_detail    rm     
H> ;                                      absolute configuration established
H>                                        by the structure determination of
H>                                        a compound containing a chiral
H>                                        reference molecule of known
H>                                        absolute configuration
H> ;
H>                                  ad     
H> ;                                      absolute configuration established
H>                                        by anomalous dispersion effects in
H>                                        diffraction measurements on the
H>                                        crystal
H> ;
H>                                  rmad   
H> ;                                      absolute configuration established
H>                                        by the structure determination of a
H>                                        compound containing a chiral
H>                                        reference molecule of known
H>                                        absolute configuration and
H>                                        confirmed by anomalous dispersion
H>                                        effects in diffraction
H>                                        measurements on the crystal.
H> ;
H>                                  syn   
H> ;                                      absolute configuration has not been
H>                                        established by anomalous dispersion
H>                                        effects in diffraction measurements
H>                                        on the crystal. The enantiomer has
H>                                        been assigned by reference to an
H>                                        unchanging chiral centre in the
H>                                        synthetic procedure.
H> ;
H>                                  unk   
H> ;                                      absolute configuration is unknown
H>                                        there being no firm chemical
H>                                        evidence for its assignment to 
H>                                        hand and it having not been
H>                                        established by anomalous
H>                                        dispersion effects in diffraction
H>                                        measurements on the crystal. An
H>                                        arbitrary choice of enantiomer
H>                                        has been made.
H> ;
H>                                  .      'inapplicable'
H>     _definition
H> ;             In 'Enantiomers, Racemates, and Resolutions' by Jean-Jacques
H>               Andre Collet and Samuel H. Wilen: John Wiley & Sons, New
H>               York, it is stated that "the absolute configuration of a
H>               chiral substance is known when an enantiomeric structure
H>               can be assigned to an optically active sample of a given
H>               sign." It is thus always recommended to report the optical
H>               activity in solution of the molecule(s) using
H>               _chemical_optical_rotation.
H> 
H>               Sufficient conditions for the assignment of
H>               _chemical_absolute_configuration are as follows, in which
H>               Set_E is defined to be the set of noncentrosymmetric crystal
H>               classes {1, 2, 3, 4, 6, 222, 32, 422, 622, 23, 432}, x(u) is
H>               the value of the Flack(1983) parameter as given by
H>               _refine_ls_abs_structure_Flack, NEAR = 1.6 and FAR = 5.0.
H> 
H>               If the crystal class is NOT in Set_E then
H>               _chemical_absolute_configuration is 'inapplicable' and must
H>               take the value '.' if present.
H> 
H>               For 'rm' to be valid the crystal class must be in Set_E and
H>               the source of the chiral reference substance of known
H>               absolute configuration must be reported.
H> 
H>               For 'ad' to be valid the crystal class must be in Set_E and
H>               |x/u| < NEAR and |(1-x)/u| > FAR.
H> 
H>               For 'rmad' to be valid the conditions of both 'rm' and 'ad'
H>               must be fulfilled.
H> 
H>               For 'syn' to be valid the crystal class must be in Set_E and
H>               the chiral centre used to assign the enantiomer must be
H>               reported.
H> 
H>                For 'unk' to be valid the crystal class must be in Set_E.
H> 
H>                Ref: Flack, H. D. (1983). Acta Cryst. A39, 876-881.
H> ;

D78.4 Mailing procedures
------------------------
D> 	In answer to your questions:
D> 
D> 1. I am in favour of moving to discussion lists, particularly if it frees
D> up some of Brian's time.
D> 
D> 2. Contributors to the list should be controlled.  People wishing to join
D> as contributors (observers) should be approved by Comcifs chair (or
D> deputy, i.e.  Brian) to ensure that we keep the discussion under control. 
D> This would probably only be necessary for Comcifs itself.  Groups working
D> under comcifs could set their own rules.  In general, the more people
D> contributing to the technical discussions on a particular dictionary, the
D> better. 
D> 
D> 3. The discussions should be visible to anyone.  We have nothing to
D> hide, and anyone who finds the discussions interesting and wished to
D> contribute could contact Brian or myself and get their name added to the
D> list of comcifs observers.
D> 
D> 4. Each group should decide how it wants to operate, but it would be a
D> good idea if this service were made available to any Comcifs group that
D> wished to use this scheme.
D> 
D> 5. The digest option looks useful. 

It is proposed to establish a discussion list for the members of the core
Dictionary Management Group (see D80.2) to test the effectiveness of this
approach.

D79.1 COMCIFS Report to the IUCr Executive
-------------------------------------------
H> > IUCr for approval following each IUCr Congress.  Voting members will be
H> 
H> would be better as:
H> 
H> IUCr for approval during or following each IUCr Congress.  Voting
H> members will be
H> 
H> > Comcifs business.  They will receive all Comcifs mailings are are free
H> 
H> should be:
H> 
H> Comcifs business.  They will receive all Comcifs mailings and are free
H> 
H> The proposals for a new strucuture for comcifs do not come clean on the
H> relationship between the DDL maintainers and Comcifs. This needs
H> elucidating before the proposals are sent to the executive.
=============================================================================

New(ish) topics
===============

D80.1 Software validation of enumeration ranges for real-number values
----------------------------------------------------------------------
(Follows from (78)D77.2.)
H> D>      The answer to Howard's problem with the enumeration range is for
H> D> the writers of software to recognise that a number given as 1.3(2) does
H> D> lie in the enumeration range 0-->1 within the experimental uncertainty.
H> D> If this interpretation is allowed, there is no need to change the
H> D> enumeration range to something that we are not yet able to handle
H> D> properly.  This convention could then be applied to any experimentally
H> D> determined number that was given with its uncertainty.  An author who
H> D> failed to report the uncertainty would then risk having the cif 
H> D> rejected (and quite rightly too).
H> 
H>   A very sensible suggestion which I can accept entirely. The definition
H> of _type numb should be extended to allow for this interpretation.


D80.2 Core Dictionary Maintenance Group
---------------------------------------
In accordance with the proposals of circular 79, David has established a
maintenance group for the core dictionary, which will take over the further
discussion and management of changes to the core dictionary. The initial
membership will be 
          I.D. Brown
          S.R. Hall
          A.L. Spek (U. Utrecht)
          P.R. Edgington
          B. McMahon
The group will conduct discussions through a mailing list with public
archives visible on the web. This will be set up early in the New Year, and 
detailed technical discussions of individual core definitions will no longer
be carried in this series of general COMCIFS circulars. This will include
the current series of discussions on absolute structure (above), and George
Sheldrick's suggestions (below) for a new way to describe the extent of
measured diffraction data sets. I hope that the next circular will have full
details of the working arrangements for this group.

D80.3 Extent of measured data
-----------------------------
G> I agree with Howard's comments that the treatment of Friedel opposites
G> should be clarified.  I would also like to repeat a further suggestion
G> that I made in D71.1 (discussing the Acta 97 instructions for authors)
G> that it is illogical to define: 
G> 
G> data_diffrn_measured_fraction_theta_full
G> data_diffrn_measured_fraction_theta_max
G> 
G> and
G> 
G> data_diffrn_reflns_theta_full
G> 
G> using the Bragg angle 'theta' rather than the resolution in Angstroms
G> (which is how a protein crystallographer would have defined it). 
G> Whereas the resolution gives an immediate indication of the extent of
G> the data (and hence the data to parameter ratio), theta depends on the
G> wavelength, which with increasing use of synchrotron data may be
G> different for different experiments.  Indeed for Laue data the
G> wavelength is different for each reflection and so 'theta' is useless
G> whereas 'resolution' would still be sensible.
G> 
G> Will COMCIFS be allowed to preview the 1998 Acta Cryst. instructions for
G> authors ?!  It might help to prevent accidents.

The 1998 Notes for Sections A and C and JSR have already been posted on the
web. As far as I am aware, there are no substantial differences from 1997.
The Editorial floor have previously been memo'd to distribute advance
copies to a number of software developers (of whom George is one), and I
shall check whether that has been done. The idea of a formal liaison between
the editorial process and COMCIFS for an early posting of the draft Notes
seems a good one, and I shall look into setting up a mechanism to ensure
that this is done in future.

D80.4 International Tables Volume G
-----------------------------------
A brief reminder to COMCIFS members who are also contributing materials to
International Tables Volume G that the 1 December deadline for initial
manuscript submission has now passed. While the editors are not without sin,
and should not therefore cast stones, I remind you that the full and
complete documentation of CIF that this volume will endeavour to provide
forms an essential element in the success of the CIF project as a whole.
Please redouble your efforts to submit your material at the earliest
opportunity.

With seasonal greetings,

Brian