Do you have a pseudotranslation in the lower spacegroup? (If I could do
symmetry in my head, I could probably work it out).
If so, then it might be worth looking at the following excellent paper...
http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=2394827
Surprises and pitfalls arising from (pseudo)symmetry
Peter H. Zwart,a* Ralf W. Grosse-Kunstleve,b Andrey A. Lebedev,c Garib
N. Murshudov,c and Paul D. Adamsab
Michael Karl Uhl wrote:
Hello again,
we looked at the graphs produced by truncate in the monoclinic SG -->
that's why we thought about twinning. The cumulative intensity
distribution does not fit to a sigmoid (indicates twinning) curve but
shows a strange distribution for the centric reflections. If I look at
the acentric and centric moments there is no straight line (particularly
for the centric reflections), the distribution is more or less something
in between twinned and not twinned.
If we index the dataset triclinic and look at the truncate graphs again,
we can find the same distribution for the acentric reflections (no
straight line). The observed cumulative intensity distribution fits very
well to the theoretic one.
Thanks for the fast answer and the help,
Michael.
Eleanor Dodson schrieb:
I dont think that is a twinning law, and that SG and cell dimensions
are unlikely to be twinned.
It could be really P1 I suppose with a twin operator -h,k,-l - ie the
SYM op is actually a twin operator..
But twinning usually shows up in the intensity statistic plots - have
you looked at the TRUNCATe plots (Import scaled task if you are
reading scalepack output to convert to mtz file)
Eleanor
Michael Karl Uhl wrote:
Hello everyone,
I have collected a dataset with a resolution of 1.5 A. The indexing with
denzo went very well with the monoclinic cell (62.3120 84.2860 72.8720
90.0000 111.9620 90.0000).
After many trials to determine the structure with MR as well as SAD I
thought that my dataset might be twinned. I reindexed the dataset
triclinic and used detwin to analyse it. The found twinning law
corresponds to "-h,-k,l". But nevertheless it was not possible for me to
determine the structure.
I was thinking about a possible twin law in P21 and tried refmac5 with
the keyword "twin" and the monoclinic dataset. Refmac5 found a twinning
law of "-h, -k, l+h" with a twinning fraction of about 20%. Detwin was
not able to deal with these data and gave following error:
*** FATAL ERROR ***
One of Twin related reflection indices -15 0 52 exceed maximum
allowed values ( 43 59 51)
Increase PARAMETER MAXIND and recompile.
Is this twinning law correct? For me it looks more like a cell
transformation than a twinnnig law.
Thanks for any help.
Best wishes, Michael
--
Mag. Michael Karl Uhl
Research Centre Applied Biocatalyses
c/o Department of Molecular Biosciences - Structural Biology
Karl-Franzens-University Graz
Humboldtstraße 50/3th floor/room 186
8010 Graz, Austria
Phone: +43 316 380 5466
Fax: +43 316 380 9897
[EMAIL PROTECTED]
http://www.Applied-Biocat.at/
http://strubi.uni-graz.at/
------------------------------------------------------------------------
------------------------------------------------------------------------
------------------------------------------------------------------------
------------------------------------------------------------------------
------------------------------------------------------------------------
------------------------------------------------------------------------
