Latest Crysfire powder indexing suite for proteins and other large cell data

Lachlan Cranswick l.cranswick at
Sat Dec 16 23:59:34 EST 2000

Latest Crysfire powder indexing suite for proteins and other large cell data

There is a new version of the Crysfire for DOS (and Windows) powder
indexing suite by Robin Shirley.  This now includes a program
called "unscale"

An tutorial using this on a large celled inorganic that 
does not solve using the raw peak find data (collected on a Philips 
1050/1710 medium resolution (173 mm goniometer radius) laboratory 
powder X-ray diffractometer - not a synchrotron) is at:

More information on "rescaling" and "unscaling" on large cells
(as well educated guesswork on volume ranges that the various 
indexing programs have been optimised for) is at:

Rescaling was the method used by Bob von Dreele and co-workers
on synchrotron powder data to help index proteins: - e.g.
    a         b        c       al     be      ga
  81.2780  81.2780  73.0390  90.000  90.000 120.000

   (R. B. Von Dreele, P. W. Stephens, G. D. Smith and R. H. Blessing, 
   "The first protein crystal structure determined from high-resolution 
   X-ray powder diffraction data: a variant of T3R3 human insulin-zinc 
   complex produced by grinding", Acta Cryst. (2000). D56, 1549-1553.)


Crysfire comes with and links seemlessly into:

       Treor90 (Per-Eric Werner) 
       Ito12 (Jan Visser) 
       Dicvol91 (D. Louër) 
       Taup (Daniel Taupin) (also called Powder - 
            Daniel Taupin's Index-Permutation program) 
       Kohl (Franz Kohlbeck) 
       FJZN (Jan Visser & R. Shirley) 
       Lzon (Robin Shirley, Daniel Louer & Jan Visser) 
       LOSHFZRF (Daniel Louer & Robin Shirley)

Summary files are generated with one solution per line making
it easy to scope the solution space of the cells:

Crysfire also includes the extra programs:

 xf2crys - program for importing "Coehlo and Cheary's" XFIT Peak Profile 
    files into CRYSFIRE CDT format.   (default filename xfit.txt) 
 wf2crys - program for importing "Stefan Krumm's" Winfit Peak Profile 
    files into CRYSFIRE CDT format. (default filename winfit.dat) 

 udi2crys - program for Philips UDI peak find data files into CRYSFIRE 
    CDT format. 

 MERGESUM Program: Merges the CRYSFIRE summary files for different 
   dataset names.


Crysfire can be downloaded from:


General tutorials on installing and running Crysfire:


>From a large celled Inorganic:

Top result on trying to index on un-rescaled data 
  (no true solution is found)

   I20  Merit    Volume   V/V1 BL IndexProg    Date     Time    Pedig
   20   35.15   1740.311  1.00 I KOHLv6.20i   27Sep00 18:33:15     27
        a        b        c     alpha    beta   gamma  
     16.9217  23.9262   4.2992  90.000  91.084  90.000

Top result with Indexing on "rescaled" data - factor of .3 - 

   I20  Merit    Volume   V/V1 BL IndexProg    Date     Time    Pedig
   20   40.78    369.668  1.00 F TAUPv3.2c    16Dec00 17:29:52 Cub__1
        a        b        c     alpha    beta   gamma
      7.1769   7.1769   7.1769  90.000  90.000  90.000

After unscaling using "unscale" - factor of .3 - to give the real 
result (confirmed originally using single crystal data)

   I20  Merit    Volume   V/V1 BL IndexProg    Date     Time    Pedig
   20   40.78  13691.390  1.00 F TAUPv3.2c    16Dec00 17:29:52 Cub__1
        a        b        c     alpha    beta   gamma 
     23.9230  23.9230  23.9230  90.000  90.000  90.000


For graphically evaluating Crysfire results: the Chekcell for Windows
program by Jean Laugier and Bernard Bochu can be very convenient and 
allows selection of possible true cells using the principle of 
"parsimony of extra reflections" by allow manual and automatic 
Best Cell solution by going through all applicable cell/spacegroup 



There are also bleeding_edge_beta_versions of the latest Chekcell
(latest beta version is quite well tested) which now has:

   LePage for Windows (from original Fortran code by A.L.Spek 
    and A. Meetsma - )
   for obtaining the reduced cell and searching for sub and super cells.

   GUI based cell transformation (Hex <-> Rhom, Obverse <-> Reverse, 
   manually define the transformation matrix, etc);
   HKL to peak assignment tolerance wizard, 

   Improved Truecell (potentially superceded by the more effecient LePage) 

   Ability to add and save manually inserted, transformed, Le page and 
   Truecell found cells to the summary list.
   Sorting of the summary list as well as ellimination of unwanted cell
    types (e.g., elliminating a few 1000 triclinic trial cells from a
    summary list).  Saving a new summary list.

   Other convenience features.
Beta versions of Chekcell are located in the  bleeding_edge_beta_versions 
  subdirectory of the above download sites.

The above Chekcell features help get around the bias in powder 
indexing programs that favour low volume low symmetry cells.

There is also a pre-release stand alone version of Le Page for Windows 
in the above mentioned bleeding_edge_beta_versions sub-directory.


Using LePage for Windows can help get around the problem described by
Buerger ('Reduced Cells', M.J. Buerger, (Zeitschift fur Kristallographie, 
BD 109, S. 42-60 (1957)) where most powder indexing programs use
the Delaunay reduction which does not reliably give the "reduced cell"

  Relevant extracts at:

  "Several authors have erroneously stated that the endpoint of the 
  Delaunay reduction is the reduced cell, although Delaunay himself was 
  well ware of the literature on reduced cells and avoided using the 
  term in connection with his reduction"


  "A short time ago Ito presented a method by which powder photographs 
  of crystals of any symmetry can be indexed.  His method indexes the 
  reflections on the basis of an arbitrary primitive cell of the lattice. 
  In order to interpret the results, this cell must be transformed into 
  some easily recognised standard cell. Ito chose, as standard, one of 
  four cells provided by the Delaunay reduction, specificially that cell 
  which is based on the three shortest axes of the four homogenious axes 
  corresponding to Delaunay reduced form. In two places, Ito referrred 
  to this cell as the "reduced cel", and this practise has been copied by 
  Henry and Lonsdale and by Donnay. This is an unfortunate slip, since 
  the term "reduced cell" was already well established, and had been 
  defined in a different way."



Lachlan M. D. Cranswick

Collaborative Computational Project No 14 (CCP14)
    for Single Crystal and Powder Diffraction
Daresbury Laboratory, Warrington, WA4 4AD U.K
Tel: +44-1925-603703  Fax: +44-1925-603124
E-mail: l.cranswick at  Ext: 3703  Room C14

More information about the Xtal-log mailing list