Re: [ccp4bb]: A simple question of resolution

[Norbert Straeter <strater@chemie.fu-berlin.de>]

***  For details on how to be removed from this list visit the  ***
***          CCP4 home page http://www.ccp4.ac.uk         ***



> You have to report the data You used for structure solution and
> refinement.
> Since this is 'common' scientific ethic You don't need a reference.
>
Well, that is correct but it implies that you have used data only to a
resolution where you still have a signal. It is easily possible to
collect data to 1.8 Ang. when you crystal diffracts just to 2.5.
This issue has been discussed  some time ago on this bulletin board and
you can find the discussion at
http://www.cryst.bbk.ac.uk/~ubcg09j/ccp4arc/bb1998/threads.html
looking for threads "weak reflections" and "Reflections + Geometry"

My contribution at that time was:
>Even if it does not do any harm to include very weak high resolution
>reflections into the refinement, it still is misleading to people if the
>structure is later reported as being refined to say 2.4 A resolution and
>it actually is a 2.8 A structure. The maximum resolution is also regarded
>as one parameter for the accuracy of the structure.
>The sigmas seem to be quite reliable with e.g. Denzo/Scalepack but I know
>other programs where they can be significantly underestimated (as judged
>by the I/sigma(I) of systematically absent reflections, for example).

I am still looking for I/sigma(I) > 2 and Rcryst < 30 but there may
reasons to include weaker data since they apparently do no harm in maximum
likelihood refinement. But than it makes sense to use independent methods
to assess the resolution such as the ones mentioned by G. Kleywegt.

-Norbert

>
>
> Robert Lucas wrote:
>
> > ***  For details on how to be removed from this list visit the  ***
> > ***          CCP4 home page http://www.ccp4.ac.uk         ***
> >
> > Since this seems to be a ripe time for debates I figure I'll
> > throw my fuel onto the fire...
> >
> > My question is a simple one, how do you determine what resolution
> > to report your structure at?
> >
> > I'm not lucky enough enough to be in a situation with multiple complete
> > MAD datasets that were solved used an ingenious program while I was
> > drinking coffee. Instead I don't like coffee and my data is incomplete
> > and gets quite sparse at high resolution.
> >
> > I'm interested in what is a proper method for reporting resolution in
> > the worst/poor case scenarios, any references to papers on the topic
> > would also be greatly appreciated.
> >
> > Thanks,
> > Robert
> >
> >      /)    Department of Molecular Biology & Biochemistry    (\
> >     / )University of California Irvine, Irvine, CA 92697-3900( \
> >   _( (    Phone: (949) 824-1933, FAX (949) 824-1954           ) )_
> > (((\ \                   Email rlucas@uci.edu                 / /)))
> > (\\\\ \_/ /     http://www.structure.uci.edu/~rlucas     \ \_/ ////)
> >  \       /                                                \       /
> >   \    _/       If you're not part of the solution,        \_    /
> >   /   /           you're part of the precipitate.           \   \
> >
> >     ***Dissertation Defense May 7th at 2PM in UCI's Tamkin F114***
>
> --
> Dr. Christopher Lehmann
>
> Center for Advanced             fon:    ++(301)-738-6126
> Research in Biotechnology       fax:    ++(301)-738-6255
> 9600 Gudelsky Drive             e-mail: lehmann@umbi.umd.edu
> Rockville, MD 20850, USA
>
> web: http://www.geocities.com/dr_christopher_lehmann/
>
>
>