[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [ccp4bb]: Odd space group
*** For details on how to be removed from this list visit the ***
*** CCP4 home page http://www.dl.ac.uk/CCP/CCP4/main.html ***
On Mon, 15 May 2000, Hal Lewis wrote:
> I'm having trouble with a MAD dataset I've collected. The protein appears
> to index well in P212121 (i.e. predictions match spots well) with the cell
> being 57.6 68.5 125.4 90.000 90.000 90.000. However the Rmerge of the
> data is high due to apparent problems with the space group I've chosen. The
> Scalepack output shows many examples of:
>
> 0 17 7 150.0 154.5 3.93
> 0 -17 -7 a 37 9.8 320.4 100 27.1
> 0 -17 -7 a 143 0.3 60.9 0 18.8
> 0 17 -7 a 37 11.8 352.5 100 25.4
> 0 17 -7 a 144 -0.3 54.3 0 19.3
>
> 0 18 7 90.4 936.7 3.72
> 0 -18 -7 a 36 -0.7 576.3 0 82.8
> 0 -18 -7 a 142 9.6 1596.1 100 109.5
> 0 18 -7 a 37 0.7 640.0 0 84.9
> 0 18 -7 a 143 7.8 1530.5 100 123.8
>
> 4 20 30 108.3 1042.7 2.61
> 4 -20 -30 a+ 32 0.3 1675.2 0 97.9
> -4 20 -30 a+ 148 -6.3 543.6 100 95.1
> -4 -20 -30 a- 147 -5.8 628.9 100 95.1
> 4 20 -30 a- 33 -0.3 1616.2 0 128.2
>
> 7 1 13 135.0 1474.9 6.24
> 7 1 13 f+ 135 0.1 2059.8 0 129.2
> -7 -1 13 f+ 46 -5.2 1026.7 98 134.7
> 7 -1 -13 f+ 55 -5.4 1033.4 99 122.3
> -7 1 -13 f+ 126 0.4 2103.2 0 134.5
> -7 -1 -13 f- 126 0.0 2036.4 0 155.1
> 7 1 -13 f- 55 -5.7 970.8 100 123.4
> -7 1 13 f- 46 -6.1 897.2 100 122.5
> 7 -1 13 f- 135 -0.4 1983.4 0 124.8
>
> etc. It appears that symmetry-related reflections aren't truly
> symmetry-related (e.g. I have the wrong space group). I've even tried
> indexing in P1 with the same result. Any idea what's going on here? I've
> collected the data twice (different crystals) on different dates to try to
> eliminate machine or crystal problems. Thanks for any help you can give me,
>
If you work through the example reflections you've given, it appears that
a better symmetry than orthorhombic is monoclinic with b* unique {e.g. for
the last set of reflections, you can split them up into two subsets, i.e.
(7 1 13), (-7 1 -13), (-7 -1 -13), (7 -1 13) and (-7 -1 13), (7 -1 -13),
(7 1 -13), (-7 1 13)}; it is not uncommon for monoclinic cells to have a
beta angle close to 90o, but the metric symmetry does not necessarily
reflect the true symmetry of the crystal system (but it should always be
at least as high).
As an example, I have collected data on a well diffracting monoclinic
crystal and indexed and refined the cell parameters as triclinic; the beta
angle refined to a better 90o (i.e. closer and with smaller esd) than the
alpha and gamma; I'm sure that most experienced crystallographers who have
collected data on more than a handful of crystals will have similar
stories.
As for your last point about the data not merging in triclinic, are you
sure that you're retaining the triclinic symmetry all the way through your
processing, and not changing back to orthorhombic at some point?
Harry
--
Dr Harry Powell, MRC Laboratory of Molecular Biology, MRC Centre, Hills
Road, Cambridge, CB2 2QH