People dont read the CCP4 documentation on twinning! Grrr
PG P3 can have 3 twinning operators; and these are:
k,h,-l ( or symm equiv) - if this is a crystallographic operator the
PG becomes P321
-h,-k,l (or symm equiv) - if this is a crystallographic operator the PG
becomes P6
-k,-h,-l (or symm equiv if this is a crystallographic operator the PG
becomes P31 2
The second moment test is not too badly affected if you get the PG wrong
( some centric reflections are flagged as acentric, but these are
usually a small % of the total)
Neither is the l test, but this is easily disturbed by problems with
the data
However the H-test, or the Britten test and some others look at
correlations between possibly twinned intensities, and there if you have
the wrong point group, they can be very misleading..
From the information you have provided I would guess the PG is P321 but
I need the TRUNCATE plots to be happy about saying that; they give some
feeling for data quality.
Eleanor
Yingjie Peng wrote:
Hi,
I tried to processed it as P321. It seemed that it might be right. The
Rmerge increased
just a little. Then I used phenix.xtriage and sfcheck to check it. The
results are as following:
phenix.xtriage:
Twinning and intensity statistics summary (acentric data):
Statistics independent of twin laws
- <I^2>/<I>^2 : 2.084
- <F>^2/<F^2> : 0.827
- <|E^2-1|> : 0.666
- <|L|>, <L^2>: 0.400, 0.227
Multivariate Z score L-test: 9.082
The multivariate Z score is a quality measure of the given
spread in intensities. Good to reasonable data are expected
to have a Z score lower than 3.5.
Large values can indicate twinning, but small values do not
necessarily exclude it.
Statistics depending on twin laws
-----------------------------------------------------------------
| Operator | type | R obs. | Britton alpha | H alpha | ML alpha |
-----------------------------------------------------------------
| -h,-k,l | M | 0.461 | 0.102 | 0.065 | 0.022 |
-----------------------------------------------------------------
Patterson analyses
- Largest peak height : 4.288
(corresponding p value : 0.98768)
The largest off-origin peak in the Patterson function is 4.29% of the
height of the origin peak. No significant pseudotranslation is detected.
The results of the L-test indicate that the intensity statistics
are significantly different than is expected from good to reasonable,
untwinned data.
As there are twin laws possible given the crystal symmetry, twinning could
be the reason for the departure of the intensity statistics from normality.
It might be worthwhile carrying out refinement with a twin specific target
function.
sfcheck:
Pseudo-translation is not detected.
Minimal estimated error : 0.0864
Perfect twinning test <I^2>/<I>^2 : 2.0191
Partial Twinning test:
-h,-k,+l
Polar angles: 0.00 0.00 180.00
Alpha(twin fraction),Npair,Ior,Tol : 0.148 118812 2 0.000
Then, what should I do? I did not deal with any twinning dataset. Any
comments and suggestions
will be greatly appreciated. Thanks!
Yingjie
On Thu, Oct 16, 2008 at 12:24 AM, Eleanor Dodson <[EMAIL PROTECTED]>wrote:
I should have said - most likely explanation is point group is reall P321
Eleanor
2008/10/15, Eleanor Dodson <[EMAIL PROTECTED]>:
I cant follow this very well.
Try SFCHECK as well which will do the same tests and give a differently
formatted output..
or TRUNCATE which gives you plots of these stats v resolution..
<I^2>/<I>^2 : 2.351 This is higher than the expected value of 2 for
untwinned data. (1.5 for perfectly twinned data)
However it can be distorted by non-crystallographic translation, but you
dont seem to have that..
Or by experimental errors and you need to inspect it in resolution
ranges
to detect that - assuming your low res data is more accurate than the
high
res.
Eleanor
Yingjie Peng wrote:
Dear guys,
I have collected a dataset with the sg as P31. I ran pehnix.xtriage to
analyse the data with
following result:
Twinning and intensity statistics summary (acentric data):
Statistics independent of twin laws
- <I^2>/<I>^2 : 2.351
- <F>^2/<F^2> : 0.788
- <|E^2-1|> : 0.766
- <|L|>, <L^2>: 0.446, 0.270
Multivariate Z score L-test: 3.358
The multivariate Z score is a quality measure of the given
spread in intensities. Good to reasonable data are expected
to have a Z score lower than 3.5.
Large values can indicate twinning, but small values do not
necessarily exclude it.
Statistics depending on twin laws
------------------------------------------------------------------
| Operator | type | R obs. | Britton alpha | H alpha | ML alpha |
------------------------------------------------------------------
| -h,-k,l | M | 0.460 | 0.043 | 0.039 | 0.022 |
| h,-h-k,-l | M | 0.054 | 0.423 | 0.459 | 0.478 |
| -k,-h,-l | M | 0.476 | 0.042 | 0.043 | 0.022 |
------------------------------------------------------------------
Patterson analyses
- Largest peak height : 4.693
(corresponding p value : 0.95672)
The largest off-origin peak in the Patterson function is 4.69% of the
height of the origin peak. No significant pseudotranslation is detected.
The results of the L-test indicate that the intensity statistics
behave as expected. No twinning is suspected.
Even though no twinning is suspected, it might be worthwhile carrying
out
a refinement using a dedicated twin target anyway, as twinned structures
with
low twin fractions are difficult to distinguish from non-twinned
structures.
The correlation between the intensities related by the twin law
h,-h-k,-l
with an
estimated twin fraction of 0.42 %
is most likely due to an NCS axis parallel to the twin axis. This can be
verified by
supplying calculated data as well.
Is it perfect twinning or partial twinning? I am supposed to do MR with
this
dataset.
What should I do next with this dataset? Thanks very much.
Best regards,
Yingjie
Yingjie PENG, Ph.D. student
Structural Biology Group
Shanghai Institute of Biochemistry and Cell Biology (SIBCB)
Shanghai Institute of Biological Sciences (SIBS)
Chinese Academy of Sciences (CAS)
320 Yue Yang Road, Shanghai 200031
P. R. China
86-21-54921117
Email: [EMAIL PROTECTED]
