Maybe we should call a spade a spade here: In any case after a modification  - even a legitimate one - of FOBS the result
should not be called FOBS anymore (F-twin1, F-twin2, whatever the mod reason). I am sure this can be fixed, but only Garib
knows what else might break if it is done. How exactly the ML map coefficients are calculated in such cases is also
not irrelevant, but in any case, contamination of 'true' FOBS data will affect data mining/re-refinement efforts.
Eventually, I guess, FOBSes as sole primary data will fade away and deposition of unmerged I may become practice.
After all the processing from individual reflections to F, how much 'OBS' is left in the FOBS? Extending the same argument
to IOBS, it seems we are almost back to image deposition....

Best, BR

-----Original Message-----
From: CCP4 bulletin board [mailto:[log in to unmask]] On Behalf Of Randy Read
Sent: Dienstag, 6. Oktober 2015 09:12
To: [log in to unmask]
Subject: Re: [ccp4bb] Relationship between Fobs/SIGFobs columns of REFMAC input and output MTZ files

Just to reinforce what Jan said: if your crystal is twinned, then the modification that Refmac carries out on what it (for some reason only known to Garib) still calls FOBS and SIGFOBS in the output MTZ is pretty extensive.

While we’re on the topic of deposition, please deposit the original IOBS and SIGIOBS, either by itself or together with the FOBS and SIGFOBS used for refinement!  Otherwise, it’s impossible to reverse the transformations made to turn I into F.

Best wishes,

Randy Read

-----
Randy J. Read
Department of Haematology, University of Cambridge
Cambridge Institute for Medical Research    Tel: +44 1223 336500
Wellcome Trust/MRC Building                         Fax: +44 1223 336827
Hills Road                                                            E-mail: [log in to unmask]
Cambridge CB2 0XY, U.K.                               www-structmed.cimr.cam.ac.uk

On 6 Oct 2015, at 06:36, Jan Dohnalek <[log in to unmask]> wrote:

You definitely should not use the Refmac output Fobs and Sigfs, those are modified. Deposition should be done with original Fobs and SIGFs and calculated Fc and phases.
J. Dohnalek


On Mon, Oct 5, 2015 at 11:31 PM, wtempel <[log in to unmask]> wrote:
Hello,
I noticed that Fobs and SIGFobs values differ between REFMAC HKLIN and HKLOUT, respectively. What is their relationship?
I calculated Rfactors using either HKLIN or HKLOUT for one example structure, and observed a difference of > 0.1%-points. Does this difference render HKLOUT unsuitable for deposition of "structure factors", even as it is HKLOUT, not HKLIN, that includes map coefficients?
Thank you for your insights.
Wolfram Tempel



--
Jan Dohnalek, Ph.D
Institute of Biotechnology
Academy of Sciences of the Czech Republic Nad Safinou II 366
252 42 Vestec near Prague
Czech Republic

Tel:   +420 226 201 571
        +420 226 201 570