David, Thanks for providing your data. my CC1/2 is often about 0.6, 0.7 since I cut at I/sigmaI =2.
Tim, you are right. I am conservative on this since I am still a new comer in this field. Charles On Thu, Sep 4, 2014 at 2:36 PM, Tim Gruene <[email protected]> wrote: > Dear Charles, > > I would ignore (high resolution) R-values for justifying the rsolution > cut-off. If you are talkig about 100s %, I have published data with > Rmeas > 500% in the highest resolution shell. > > You are quite save cutting the resolution where I/sigI >= 2.0, by todays > knowledge this is actually conservative. Another good criterion would be > the combination of CC1/2 = 40% AND I/sigI > 1.0 if you want to extend > the resolution a litte. If I understand correctly, this latter combined > criterion indicates that the standard deviations are estimate correctly. > This is important so that downstream programs work, notably maximum > likelyhood based refinement programs like refmac or buster. > > Best, > Tim > > > On 09/04/2014 07:25 PM, CPMAS Chen wrote: > > Hi, All CCP4BB Users, > > > > I have quite some data sets(~15) collected at different beam intensities, > > the individual dataset can diffract to ~3.8 A @I/dI=2. If I combine > them, > > with AIMLESS in CCP4, the resolution can be extended to ~3A@I/dI=2. But, > > the Rmerge or Rpim is way high, in 10s or even 100s. Would this high R > > simply due to the intensity difference among these datasets? Or simply > say, > > can I trust/use the combined/merged dataset diffracting to 3A? > > > > Thanks! > > > > Charles > > > > -- > Dr Tim Gruene > Institut fuer anorganische Chemie > Tammannstr. 4 > D-37077 Goettingen > > GPG Key ID = A46BEE1A > > -- *************************************************** Charles Chen Research Associate University of Pittsburgh School of Medicine Department of Anesthesiology ******************************************************
