The XDS CORRECT step will also (by default) do scaling - if I
understand correctly the same scaling as in XSCALE, but for a single
sweep and with no zero-dose.
If what you want is merging statistics, I find it helpful to write out
the data unmerged and then use pointless -c and aimless to merge the
data (with scales constant) which provides a very nice summary.
Best wishes,
Graeme
On 16 January 2013 13:19, Tim Gruene <[log in to unmask]> wrote:
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
>
> Dear Sebastiano,
>
> if the output of GO.COM produces an mtz-file you can use for phasing
> or refinement, I am very confident there is a scaling step involved.
> This might also be an explanation for the discrepancy you point out:
> With FRIEDEL'S_LAW=TRUE the statistics which affects the rejection of
> outliers would have more reflections within each group of symmetry
> related reflections and hence a greater spread which might lead to the
> rejection of some of the classes. But this is a mere guess.
>
> Cheers,
> Tim
>
> On 01/16/2013 02:03 PM, Sebastiano Pasqualato wrote:
>>
>> Thanks to Kay and Tim or the feedback.
>>
>> The reason I wanted to get statistics from the CORRECT step of XDS
>> is that I have refined a structure using the mtz output by the
>> GO.COM automatic reduction routine of SLS beamline PXIII, which
>> does not involve a scaling step (I discovered recently). I was
>> willing to have the integration statistics of the reflection file I
>> used in the refinement's high resolution bin. I will definitely
>> give xprep a try.
>>
>> Another question that raised by looking deeper into their automatic
>> procedure (thanks Meitian for the help) is that when integrating
>> with XDS CORRECT keeping the FRIEDEL'S_LAW=TRUE or =FALSE I get a
>> different number of reflections in the final mtz.
>>
>> In my case, if I run the same XDS.INP script and change only the
>> FRIEDEL'S_LAW flag, I obtain:
>>
>>
>> =TRUE: 11551 reflections
>>
>> * Resolution Range :
>>
>> 0.00043 0.11138 ( 48.225 - 2.996 A )
>>
>> * Sort Order :
>>
>> 1 2 3 0 0
>>
>> * Space group = 'P 3 2 1' (number 150)
>>
>>
>>
>> OVERALL FILE STATISTICS for resolution range 0.000 - 0.111
>> =======================
>>
>>
>> Col Sort Min Max Num % Mean Mean Resolution
>> Type Column num order Missing complete abs.
>> Low High label
>>
>> 1 ASC 0 29 0 100.00 13.7 13.7 48.22 3.00
>> H H 2 NONE 0 16 0 100.00 4.7 4.7 48.22
>> 3.00 H K 3 NONE -32 32 0 100.00 0.5 12.2
>> 48.22 3.00 H L 4 NONE 1.4 292.0 0 100.00 19.40
>> 19.40 48.22 3.00 F FP 5 NONE 0.1 4.3 0 100.00
>> 0.70 0.70 48.22 3.00 Q SIGFP 6 NONE 0.0 1.0 0
>> 100.00 0.95 0.95 48.22 3.00 I FreeRflag
>>
>>
>> No. of reflections used in FILE STATISTICS 11551
>>
>>
>> =FALSE: 11643 reflections
>>
>> * Cell Dimensions : (obsolete - refer to dataset cell dimensions
>> above)
>>
>> 100.5450 100.5450 96.4500 90.0000 90.0000 120.0000
>>
>> * Resolution Range :
>>
>> 0.00043 0.11138 ( 48.225 - 2.996 A )
>>
>> * Sort Order :
>>
>> 1 2 3 0 0
>>
>> * Space group = 'P 3 2 1' (number 150)
>>
>>
>>
>> OVERALL FILE STATISTICS for resolution range 0.000 - 0.111
>> =======================
>>
>>
>> Col Sort Min Max Num % Mean Mean Resolution
>> Type Column num order Missing complete abs.
>> Low High label
>>
>> 1 ASC 0 29 0 100.00 13.7 13.7 48.22 3.00
>> H H 2 NONE 0 16 0 100.00 4.7 4.7 48.22
>> 3.00 H K 3 NONE -32 32 0 100.00 0.5 12.2
>> 48.22 3.00 H L 4 NONE 1.4 291.9 0 100.00 19.40
>> 19.40 48.22 3.00 F FP 5 NONE 0.1 4.3 0 100.00
>> 0.67 0.67 48.22 3.00 Q SIGFP 6 NONE -13.6 13.2 69
>> 99.41 -0.01 0.69 48.22 3.00 D DANO 7 NONE 0.0
>> 5.7 69 99.41 1.13 1.13 48.22 3.00 Q SIGDANO 8
>> NONE 0 2 0 100.00 0.0 0.0 48.22 3.00
>> Y ISYM 9 NONE 0.0 1.0 0 100.00 0.95 0.95
>> 48.22 3.00 I FreeRflag
>>
>>
>> No. of reflections used in FILE STATISTICS 11643
>>
>>
>> Aren't they supposed to be the exact same number?
>>
>> Thanks, ciao, s
>>
>>
>> On Jan 16, 2013, at 1:12 PM, Tim Gruene <[log in to unmask]>
>> wrote:
>>
>> Dear Sebastiano,
>>
>> you could use xprep to get the statistics in user defined
>> resolution shells.
>>
>> Out of curiosity: Would you mind sharing why you want to do this
>> and why you don't want to use the XSCALE statistics instead? The
>> statistics are probably more meaningful after scaling, I guess.
>>
>> Best, Tim
>>
>> On 01/15/2013 04:22 PM, Sebastiano Pasqualato wrote:
>>>>>
>>>>> Hi all, I was wondering if XDS allows to change the number
>>>>> of resolution bins appearing in the table:
>>>>>
>>>>> SUBSET OF INTENSITY DATA WITH SIGNAL/NOISE >= -3.0 AS
>>>>> FUNCTION OF RESOLUTION
>>>>>
>>>>> of CORRECT.LP.
>>>>>
>>>>> Please, note that I am not referring to the table output by
>>>>> XSCALE, in which you can change the resolution bins with the
>>>>> keyword RESOLUTION_SHELLS=, but rather the table output by
>>>>> the CORRECT job of XDS.
>>>>>
>>>>> Thanks in advance, ciao, Sebastiano
>>>>>
>>>>>
>>
>>
>>
>
> - --
> - --
> Dr Tim Gruene
> Institut fuer anorganische Chemie
> Tammannstr. 4
> D-37077 Goettingen
>
> GPG Key ID = A46BEE1A
>
> -----BEGIN PGP SIGNATURE-----
> Version: GnuPG v1.4.12 (GNU/Linux)
> Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/
>
> iD8DBQFQ9qj9UxlJ7aRr7hoRAic9AJ9P8Yg7qoSwtKETrt1uVV9zmzrddgCgz3Ub
> Gd+HUldC4e4RDI7f7qzJtTw=
> =xABX
> -----END PGP SIGNATURE-----
|