Re: [ccp4bb]: Re: Resolution (Bart's reply to the Rsym thread)

[Anastassis Perrakis <perrakis@nki.nl>]

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

> It seems to me that all 1.5 A datasets, for instance, are not equal. If I have a
> crystal that really diffracts to 1.3 A and instrumental limitations prevent
> collecting past 1.5 A, the maps, resolution of atoms, etc. look considerably
> better than a more standard 1.5 A dataset and approach those seen for a 1.3 A
> dataset.

Moreover, a crystal with diffraction limit at 2.0 A in an old rotating anode with no
mirrors gives better
maps than a dataset with a diffraction limit at 2.0 A from a sooper-dooper beamline.

The first one is likely to have  Wilson B of ~5 and the other one of about ~60 ...

The fact that 'some 2.0 datasets are more equal than other 2.0 datasets' is
illustrated nicely in autotracing:
some 2.0 datasets trace in an hour and a few never do ....

As for the Rsym and I/sigI my opinion would be that as long as the cutoff criteria
are clearly stated
and explained at the table legend I fail to see why a referee would object. Maybe
Rsym 80% and I/sigI of 0.1
would need better justification than Ram's values though ...

I would suggest to Rams to simply illustrate to the referee that if he limits the
multiplicity the Rsym goes
down so the dataset can 'look' better while it is obviously not. My limited personal
experience is that when I/sigI drops below 1.5 (Rsym is then around 50% unless there
is a pathological problem with the crystal) the data tend to be of not much use as
judged by automated building and refinement. I had so far more than a handfull of
'external' cases though, where over-conservative cutoffs (I/sigI 2.0 and Rmerge less
than 35%) were not allowing automated model building and refinement to proceed. When
these datasets were truncated a bit less stringently (0.1 - 0.2 A better data,
10-20% more reflections) and then a complete model was build and refined in hours
with arp/warp.


    Tassos
--
NKI, Department of Molecular Carcinogenesis
Plesmanlaan 121 1066 CX Amsterdam NL
Phone +31-20-512-1951 Fax +31-20-512-1954
SMS: +31-6-28597791 WWW: http://den.nki.nl