refmac low resolution REMARK

From: Bernhard Rupp (Hofkristallrat a.D.)
Date: 2 January 2012 06:58


Dear Developers,

it seems to me that at least Refmac 5.6.0117 does not report the actual
lowest resolution used in refinement, but instead the lowest resolution of
the hkl index range generated by the default cif import script instead.

REMARK   3  DATA USED IN REFINEMENT.
REMARK   3   RESOLUTION RANGE HIGH (ANGSTROMS) :   1.99
REMARK   3   RESOLUTION RANGE LOW  (ANGSTROMS) :  38.63 <---

 The data go only to 28.66 A according to mtzdump as well as xprep

 as far as I can tell, cif2mtz does it right and reports

* Cell Dimensions : (obsolete - refer to dataset cell dimensions above)

  33.1300   57.2300   38.6500   90.0000   91.9400   90.0000

 *  Resolution Range :

   0.00122    0.25305     (     28.660 -      1.988 A ) < -----

then the script calls unique  which pads the reflections up

 * Cell Dimensions : (obsolete - refer to dataset cell dimensions above)

  33.1300   57.2300   38.6500   90.0000   91.9400   90.0000

 *  Resolution Range :

   0.00067    0.25302     (     38.628 -      1.988 A ) <-----

which then propagates to CAD and FREEFLAG into the MTZ header where it is
picked up by refmac.

OVERALL FILE STATISTICS for resolution range   0.001 -   0.253
 =======================
 Col Sort    Min    Max    Num      %     Mean     Mean   Resolution   Type
Column
 num order               Missing complete          abs.   Low    High
label

  1 ASC    -16      16      0  100.00      0.1      6.3  38.63   1.99   H
H
  2 NONE     0      28      0  100.00     10.5     10.5  38.63   1.99   H
K
  3 NONE     0      19      0  100.00      7.3      7.3  38.63   1.99   H
L
  4 NONE    0.0     9.0     0  100.00     4.50     4.50 38.63 1.99   I
FREE
  5 NONE    7.7   707.8   361   96.40   102.76   102.76  28.66   1.99   F
FP <----- this is the right number!
  6 NONE    1.6    37.3   361   96.40     8.84     8.84  28.66   1.99   Q
SIGFP


I have noticed that other PDB/sf file sets (not mine) seem to be affected as
well.....

Best regards, BR

----------
From: Ian Tickle


Hi Bernhard

So what's new?  See
http://ccp4bb.blogspot.com/2011/11/pdb-header-info-wrong.html
(17-Nov-11).

Cheers

-- Ian


On 2 January 2012 06:58, Bernhard Rupp (Hofkristallrat a.D.)

----------
From: Bernhard Rupp (Hofkristallrat a.D.) <hofkristallamt@gmail.com>
Date: 2 January 2012 23:18
To: CCP4BB@jiscmail.ac.uk


> So what's new?

Alzheimer's and late night senile dementia :-)

BR

PS: nice site!