> -----Original Message-----
> From: [log in to unmask] [mailto:[log in to unmask]]
On
> Behalf Of Eleanor Dodson
> Sent: 09 September 2008 17:11
> To: William G. Scott
> Cc: [log in to unmask]
> Subject: Re: [ccp4bb] truncate and anisotropy
>
> I dont think you need to worry much about the strong stacking
> reflections; TRUNCATE really only modifies the weakest data - and
> anything > 3Sigma is barely altered . There is a slightly wrong
estimate
> of <sigma> for that resolution shell, but in practice it seems to have
> no observable effect..
>
> Anisotropy is another question and drastically increases the number of
> weak reflections in the higher resolution shells which must be bad..
Ian
> - have you analysed any such data sets?
I must admit I've not looked specifically at the effect of anisotropy,
pseudo-translation etc on the Wilson distribution. But it should (in
principle at least) be possible to do something about this, i.e. one
could obtain the experimental PDF from the measured intensities and use
that in the Bayes integrals to get the corrected intensities, rather
than naively using the Wilson PDFs as we are doing now. One can get the
CDFs (cumulative distribution function) for the centric & acentric data
by plotting Z (normalised intensity) = I/(eS) (where e = epsilon factor,
S = mean epsilon-corrected intensity for shell) along the x axis,
against (i-.5)/n (where i is the serial no of the reflection after
sorting in increasing Z value and n is the no of centric or acentric
reflections) along y. The PDF is then just the 1st derivative of the
CDF, which can be obtained by finite differences, maybe with some
smoothing to smooth out any 'noise' due to sampling. I don't know how
well this scheme would work in practice, it would need some testing.
The question remains whether to do this for all centric data in one go &
similarly for all acentric data, or to split each type of data into
resolution shells. In the first case one is assuming that the PDF is
the same at all resolutions, which may not be appropriate say in the
case of a pseudo-translation, because usually such a translation applies
uniformly to all atoms only at low-medium resolution & at high
resolution it often breaks down, so one would expect a reversion to the
Wilson PDF at high res. How this plays out in the case of anisotropy
I've no idea (one would certainly need to derive an overall anisotropic
B factor correction), you would have to suck it & see. If you decide to
do it in resolution shells the centric data might prove to be a problem:
there may not be enough of them in each shell to give a reliable
estimate of the PDF, in which case you might say have to revert to using
all centric data in one go, while still keeping the acentric data in
shells.
-- Ian
Disclaimer
This communication is confidential and may contain privileged information intended solely for the named addressee(s). It may not be used or disclosed except for the purpose for which it has been sent. If you are not the intended recipient you must not review, use, disclose, copy, distribute or take any action in reliance upon it. If you have received this communication in error, please notify Astex Therapeutics Ltd by emailing [log in to unmask] and destroy all copies of the message and any attached documents.
Astex Therapeutics Ltd monitors, controls and protects all its messaging traffic in compliance with its corporate email policy. The Company accepts no liability or responsibility for any onward transmission or use of emails and attachments having left the Astex Therapeutics domain. Unless expressly stated, opinions in this message are those of the individual sender and not of Astex Therapeutics Ltd. The recipient should check this email and any attachments for the presence of computer viruses. Astex Therapeutics Ltd accepts no liability for damage caused by any virus transmitted by this email. E-mail is susceptible to data corruption, interception, unauthorized amendment, and tampering, Astex Therapeutics Ltd only send and receive e-mails on the basis that the Company is not liable for any such alteration or any consequences thereof.
Astex Therapeutics Ltd., Registered in England at 436 Cambridge Science Park, Cambridge CB4 0QA under number 3751674
|