Hi Nicolas,
-------- הודעה מקורית --------
מאת Nicolas Soler <nso...@mrc-lmb.cam.ac.uk>
תאריך: 23/09/2013 18:40 (GMT+02:00)
אל CCP4BB@JISCMAIL.AC.UK
נושא Re: [ccp4bb] tNCS: indexing and EP problems
You may want to make sure that all other parameters in XDS are the same as in mosflm and particularly the origin. Otherwise, how does the xds indexing/integration look? Perhaps pointless will give a clue about the relation between the xds and mosflm cells/spacegroup?
Cheers, Boaz
-------- הודעה מקורית --------
מאת Nicolas Soler <nso...@mrc-lmb.cam.ac.uk>
תאריך: 23/09/2013 18:40 (GMT+02:00)
אל CCP4BB@JISCMAIL.AC.UK
נושא Re: [ccp4bb] tNCS: indexing and EP problems
Thanks for your answers, I realized too late that my question number 2
was misleading. I wondered in fact whether shelxd or another program
could handle tNCS fine.
Otherwise for my indexing concerns, I have been trying to feed xds with
the unit cell parameters found by mosflm via the UNIT_CELL_CONSTANTS
keyword in XDS.INP but I didn't find them back in the solution list
proposed by IDXREF.LP. Has anybody had similar issues?
Cheers,
Nicolas
On 23/09/2013 15:22, Tim Gruene wrote:
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
>
> Dear Nicolas,
>
> shelxd is one answer to question 2
> (http://shelx.uni-ac.gwdg.de/SHELX/shelxd_mm_keywords.php: PATS). I am
> sure there are others.
>
> You can provide the cell from mosflm to XDS. If you want XDS to find
> the cell without this information, there are plenty of keywords to
> tweak, although in my experience it is often sufficient to find a
> decent set of images to collect spots from, play with the things you
> let XDS refine (REFINE(...)) and finding the correct ORGX and ORGY.
>
> In difficult cases I use adxv to get an estimate for the cell
> dimensions to judge whether or not XDS found the correct one.
>
> Best,
> Tim
>
> On 09/23/2013 04:04 PM, Nicolas Soler wrote:
>> Dear experts,
>>
>> I am dealing at the moment with a case involving translated NCS
>> copies of my asymmetric unit along one axis of the unit cell (3
>> clear non-origin peaks in the native Patterson).
>>
>> I could get Mosflm to find the corresponding big unit cell only
>> after restricting the "max deviation from integral hkl" parameter
>> to 0.1 and thus get the majority of the spots under prediction
>> boxes (although the cell parameters don't look very accurate
>> sometimes).
>>
>> I have 2 questions about it :
>>
>> 1) Do you know whether it would be possible to configure XDS the
>> same way, to make it find this enlarged unit cell ? (no success so
>> far...)
>>
>> 2) Are you aware of any experimental phasing program making use of
>> the information provided by the Patterson for finding the
>> substructure?
>>
>> Many thanks in advance,
>>
>> Nicolas
>>
> - --
> - --
> Dr Tim Gruene
> Institut fuer anorganische Chemie
> Tammannstr. 4
> D-37077 Goettingen
>
> GPG Key ID = A46BEE1A
>
> -----BEGIN PGP SIGNATURE-----
> Version: GnuPG v1.4.14 (GNU/Linux)
> Comment: Using GnuPG with Mozilla -http://enigmail.mozdev.org/
>
> iD8DBQFSQE6XUxlJ7aRr7hoRAnUEAJ9l0CNtIOuFytwWO1+MY7zYijcRKQCbBibF
> 0RJPhc5dGV0LJlzsRozjy6c=
> =7JHE
> -----END PGP SIGNATURE-----
was misleading. I wondered in fact whether shelxd or another program
could handle tNCS fine.
Otherwise for my indexing concerns, I have been trying to feed xds with
the unit cell parameters found by mosflm via the UNIT_CELL_CONSTANTS
keyword in XDS.INP but I didn't find them back in the solution list
proposed by IDXREF.LP. Has anybody had similar issues?
Cheers,
Nicolas
On 23/09/2013 15:22, Tim Gruene wrote:
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
>
> Dear Nicolas,
>
> shelxd is one answer to question 2
> (http://shelx.uni-ac.gwdg.de/SHELX/shelxd_mm_keywords.php: PATS). I am
> sure there are others.
>
> You can provide the cell from mosflm to XDS. If you want XDS to find
> the cell without this information, there are plenty of keywords to
> tweak, although in my experience it is often sufficient to find a
> decent set of images to collect spots from, play with the things you
> let XDS refine (REFINE(...)) and finding the correct ORGX and ORGY.
>
> In difficult cases I use adxv to get an estimate for the cell
> dimensions to judge whether or not XDS found the correct one.
>
> Best,
> Tim
>
> On 09/23/2013 04:04 PM, Nicolas Soler wrote:
>> Dear experts,
>>
>> I am dealing at the moment with a case involving translated NCS
>> copies of my asymmetric unit along one axis of the unit cell (3
>> clear non-origin peaks in the native Patterson).
>>
>> I could get Mosflm to find the corresponding big unit cell only
>> after restricting the "max deviation from integral hkl" parameter
>> to 0.1 and thus get the majority of the spots under prediction
>> boxes (although the cell parameters don't look very accurate
>> sometimes).
>>
>> I have 2 questions about it :
>>
>> 1) Do you know whether it would be possible to configure XDS the
>> same way, to make it find this enlarged unit cell ? (no success so
>> far...)
>>
>> 2) Are you aware of any experimental phasing program making use of
>> the information provided by the Patterson for finding the
>> substructure?
>>
>> Many thanks in advance,
>>
>> Nicolas
>>
> - --
> - --
> Dr Tim Gruene
> Institut fuer anorganische Chemie
> Tammannstr. 4
> D-37077 Goettingen
>
> GPG Key ID = A46BEE1A
>
> -----BEGIN PGP SIGNATURE-----
> Version: GnuPG v1.4.14 (GNU/Linux)
> Comment: Using GnuPG with Mozilla -http://enigmail.mozdev.org/
>
> iD8DBQFSQE6XUxlJ7aRr7hoRAnUEAJ9l0CNtIOuFytwWO1+MY7zYijcRKQCbBibF
> 0RJPhc5dGV0LJlzsRozjy6c=
> =7JHE
> -----END PGP SIGNATURE-----
