Validation Tests on CIF & CCDC-FDAT-data
(Related to but not equivalent to the IUCR-Checking)

VALIDATION

X-Ray crystallography has rightly the name of providing SOLID information on molecules in the crystalline state. However, the often forgotten condition to justify such a label for a particular instance of a crystal structure determination is that the underlying procedures should have been carried out adequately and correctly. Unfortunately, that is not always the case. Major causes are related to lack of experience in picking up signals indicating problems. Objective validation procedures attempt to address a large number of potential problems as a routine procedure.

An excellent paper by R.L. Harlow, detailing many of the issues involved, can be found in J. of Research of the National Institute of Standards and Technology, (1996), 101, 327-339. Troublesome Crystal Structures: Prevention, Detection, and Resolution.

Structural data provided in CIF-format may be validated in terms of consistency, missing relevant data, proper procedure and overlooked problems.

PLATON implements a large variety of such tests specified in check.def

The PLATON validation test set includes many of the official IUCR Validation Procedures in use for the validation of Acta Cryst. papers reporting single crystal structures.

This set is complemented with a more involved analysis of the data themselves with techniques available in PLATON (ADDSYM/VOIDS/Thermal Motion Analysis etc.)

It should be realized that validation has at least three faces:

Checking for omissions & obvious errors.
Refined structures should be routinely checked for these problems prior to submission to a journal.
Common problems include misplaced hydrogen atoms.
Pointing to technical issues to be discussed in some detail
This includes a discussion on issues related to pseudo-symmetry, disorder etc.
Pointing at unusual but potentially interesting structural features.
This includes interesting short intermolecular contacts.

Note: A subset of the the validation checks is applied for CCDC-FDAT structured files.

Implementation

The validation feature in PLATON may be invoked in three ways:

via the VALIDATION button on the PLATON-Opening window.
via: 'platon -u data.cif'
with instruction 'VALID' from the keyboard

In all cases, a copy of the file 'check.def' should be present in the current directory.

Alternatively, the path to check.def may be given via the environment variable 'CHECKDEF'.

Example: setenv CHECKDEF /mnt/user/check.def

The results of the validation can be found on the file '<dataname>.chk'

The preparation of a small shell script may be helpful. Below is an example of such a script that can be made available systemwide by placing it in /usr/local/bin

#!/bin/csh
ln -sf /mnt/spea/acc/check.def
platon -u $1.cif
/usr/sbin/unlink check.def
less $1.chk

The validation than runs for CIF sk1000.cif by:

chk sk1000

Example of a Validation Report


#===============================================================================
# PLATON/CHECK run versus check.def version of 08/02/00      for entry:  I      
# Data From: e60.cif - Data Type: CIF  

# CELL  1.54178  16.645  25.768   7.066   90.00   90.00   90.00   3030.70
# SpaceGroup P212121             Rep: P 21 21 21 
# MoietyFormula C34 H52 O4
#      Reported C34 H52 O4
# SumFormula C34 H52 O4   Rep: C34 H52 O4
# Mr        =    524.76[Calc],    524.76[Rep]
# Dx,gcm-3  =     1.150[Calc],     1.150[Rep]
# Z         =         4[Calc],         4[Rep]
# Mu (mm-1) =     0.568[Calc],     0.568[Rep]
# Reported   T limits: Tmin=0.940              Tmax=0.970 '\Y SCAN'      
# Calculated T limits: Tmin=0.815 Tmin'=0.797  Tmax=0.945
# Reported   Hmax= 20, Kmax= 31, Lmax=  8, Nref= 3292       , Th(max)= 70.01
# Calculated Hmax= 20, Kmax= 31, Lmax=  8, Nref= 3291( 5758), Ratio= 1.00( 0.57)
# R= 0.0463( 2929), wR2= 0.1367( 3292), S = 1.162, Npar= 347, Flack=-0.40( 4)   
#===============================================================================

>>> The Following ALERTS were generated <<<

052_ALERT A (proper) absorption correction method missing              
702_ALERT A Rep. Angle    174   Dev. from Calcd  110.00          5.00 Sigma     
702_ALERT A Rep. Angle    176   Dev. from Calcd  110.00          5.00 Sigma     
702_ALERT A Rep. Angle    177   Dev. from Calcd  109.00          5.00 Sigma     
702_ALERT A Rep. Angle    183   Dev. from Calcd  109.00          5.00 Sigma     
#===============================================================================
222_ALERT B Large Non-Solvent    H     Ueq(max)/Ueq(min) .       4.02 range    
#===============================================================================
032_ALERT C Std Uncertainty in Flack parameter too high ..       0.40          
033_ALERT C Flack Parameter value deviates from zero .....      -0.40          
057_ALERT C Correction for absorption required   RT(exp) =       1.16          
089_ALERT C Poor Data / Parameter ratio ..................       9.49          
220_ALERT C Large Non-Solvent    C     Ueq(max)/Ueq(min) .       3.23 range     
412_ALERT C Short intra XH3 .. XHn  :  H(242) ..  H(253) =       1.81 Ang.      
412_ALERT C Short intra XH3 .. XHn  :  H(251) ..  H(262) =       1.83 Ang.      
701_ALERT C Rep. Bond      61   Dev. from Calcd  0.9614          1.43 Sigma     
701_ALERT C Rep. Bond      81   Dev. from Calcd  0.9588          1.23 Sigma     
702_ALERT C Rep. Angle    167   Dev. from Calcd  109.40          1.03 Sigma     

ALERT Level Summary
  5 ALERT Level A = In General: Serious Problem
  1 ALERT Level B = Potentially Serious Problem
 10 ALERT Level C = Check & Explain
#===============================================================================

PLATON HOMEPAGE

11-Oct-2002 A.L.Spek

Validation Tests on CIF & CCDC-FDAT-data (Related to but not equivalent to the IUCR-Checking)

VALIDATION

Implementation

Example of a Validation Report

Validation Tests on CIF & CCDC-FDAT-data
(Related to but not equivalent to the IUCR-Checking)