[phenixbb] Anomalous or not?

Tim Gruene tg at shelx.uni-ac.gwdg.de
Thu Jul 10 07:47:50 PDT 2014


-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Hi Francis,

I couldn't read it properly, but I see there seems a certain
similarity between

v(I)=a*(v0(I)+b*I^2)

and

􏰇0(Ihl) = Sdfac[􏰇2(Ihl) + (Sdadd ghlhIhi)2]1/2.

and doing it twice may corrupt the sigmas. It is better to apply an
error model only once, no matter what error model.

Best,
Tim

On 07/10/2014 04:38 PM, Francis Reyes wrote:
> Hi Tim
> 
> So from the XDS docs, the estimates of the sigmas are performed by
> the CORRECT step in the following way:
> 
> "The residual scatter in intensity of symmetry-equivalent
> reflections is used to estimate their standard deviations. Here,
> the initial estimate v0(I) (obtained from the INTEGRATE step) for
> the variance of the reflection intensity I is replaced by
> v(I)=a*(v0(I)+b*I^2). The two constants a and b are chosen to
> minimize discrepancies between v(I) and the variance estimated from
> sample statistics of symmetry related reflections. Based on the
> more realistic error estimates for the intensities, outliers are
> recognized by comparison with other symmetry-equivalent
> reflections. These outliers are included in the main output file
> XDS_ASCII.HKL in which they are marked by a negative sign attached
> to the estimated standard deviations of their intensity...."
> 
> SCALA/AIMLESS have a completely different way of estimating the
> sigmas.
> 
> "After scaling, the error estimates can be improved by comparing
> the observed scatter between observations and the estimated
> standard deviation, making them equal on average. If the standard
> deviations 􏰇(Ihl) are correct, then the normalized deviations 􏰂hl
> = (Ihl - hIh0 i)/􏰇(Ihl) (where hIh0 i is averaged over all
> observations of reflection h excluding the lth observation) should
> be distributed with a mean 0.0 and stan- dard deviation 1.0. A
> simple correction to give improved error estimates is 􏰇0(Ihl) =
> Sdfac[􏰇2(Ihl) + (Sdadd ghlhIhi)2]1/2. "
> 
> 
> And are the two models for estimating the sigmas compatible?
> 
> F
> 
> 
> On Jul 10, 2014, at 9:22 AM, Tim Gruene <tg at SHELX.UNI-AC.GWDG.DE>
> wrote:
> 
>> The CORRECT step scales your data. It appears that autoxds will
>> call aimless or scala to scale your data again. This is little
>> harmful for the intensities, but corrupts the sigma values
>> because the error model is applied twice and thus it is harmful
>> for phasing using the anomalous signal because this relies on
>> proper sigma values.
> 

- -- 
- --
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 Icedove - http://www.enigmail.net/

iD8DBQFTvqeWUxlJ7aRr7hoRAhueAKDJCeALNTaLC8uoLQImWRl4KWhukQCeP4ok
LT4mM8ggc3iouehqpzxMHPw=
=oFme
-----END PGP SIGNATURE-----


More information about the phenixbb mailing list