[ccp4bb] Mean B value in refmac5

2023-11-05 Thread Qixu Cai
Dear all, I found that the "Mean B value (OVERALL, A**2)" reported by refmac5 in the head region of pdb file is 62.76. However, I calculated the average B value of all B factors from all atoms and the result is 67.7. What makes the difference? The canonical restrained refinement mode was used in r

Re: [ccp4bb] Mean B value in refmac5

2023-11-05 Thread Ian Tickle
The arithmetic mean B value from the structure as quoted everywhere is pretty meaningless anyway and 10 Ang.^2 either way is probably not significant. Let's say some waters or LYS side-chains or whatever have B = 1000 Ang.^2. That will bias the mean B upwards, but those atoms do not contribute sign

Re: [ccp4bb] About model building

2023-11-05 Thread Sam Tang
Dear all Thanks for all the input. I will definitely try out in the coming couple of days. To provide more information: - the data was checked in Xtriage and no major pathology was found. Completeness was 99.2% with multiplicity of >13. Mean I/sigma(I) 18.8, CC1/2 for outer shell 0.902. - After on

Re: [ccp4bb] About model building

2023-11-05 Thread Eleanor Dodson
Hmmm - r 50% rather suspect.. do the MR solutions from phaser and molten agree? Eleanor On Mon, 6 Nov 2023 at 05:10, Sam Tang wrote: > Dear all > > Thanks for all the input. I will definitely try out in the coming couple > of days. To provide more information: > - the data was checked in Xtriage

Re: [ccp4bb] Mean B value in refmac5

2023-11-05 Thread Eleanor Dodson
Are there some atoms with occupancies < 1.0 ? You would need to weight those Bs by occ On Sun, 5 Nov 2023 at 11:34, Ian Tickle wrote: > The arithmetic mean B value from the structure as quoted everywhere is > pretty meaningless anyway and 10 Ang.^2 either way is probably not > significant. Let's