[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [ccp4bb]: A simple question of resolution
*** For details on how to be removed from this list visit the ***
*** CCP4 home page http://www.ccp4.ac.uk ***
>
> > strategy program, you should be able to get good completeness. The main
> > refinement programs, like CNS, use a 1 or 2 sigma cutoff for refinement.
> > So if I process the data to 1sig, then I refine to 1sig.
> >
>
> Why use a cut off in refinement? This isunecessary and a Bad Thing to do
....furthermore: xplor in old days was using this appauling 2 sigma cutoff -
defaults in CNS are to use all data - if my memory serves me well.
..... let me also light up the fire with an observation:
Recently I got some 2.0 A data from Ashley Deacon for a project of his. It was a
260 residues protein and he wanted to autotrace it with descent phases. The xtal
had very low solvent content and this the 2.0 A data were for marginal for
arp/warp, it could get basically nothing.
I asked him to reprocess and include the 1.8 A data, which had very high R's
(above 50% if I recall correctly) and I/igI of ~1, for unique reflections.
The autotracing produced ~240 residues.
What am I tempted to conclude from that ?
1. If you run arp/warp, include weak high resolution data - refmac5 takes good
care.
2. If that bad data helps the refinement AND also improves the map (else
autotracing would not be getting more residues !) they DO help both to get the
refinement going AND the map so they are Good for You.
So, I would tend to be not so conservative these days and assume that sigma's and
sigmaa do my job.
What to report to the paper is another issue. If it goes to Nature it always
helps to cut the resolution below 3.0 - looks more challenging ;-))) I would vote
for Alexey's optical resolution from SFCHECK.
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