Hi Florian,
I would also like to refine f' and f'', as the data are from a selenomethionine derivatized protein. From what I can see the default in Phenix.Autosol seems to be not to keep the F+/- and Dano labels for the output mtz file? So I suppose I do need to provide the initial mtz file or is there a way to change the output options for the coefficients in Phenix Autosol?
all you need to refine f' and f'' is the experimental data Fobs+/Fobs-. phenix.refine does not use Dano (I don't even know what it is). Also you need to specify the anomalous scatterers. For details see: http://phenix-online.org/documentation/refinement.htm#anch22 http://phenix-online.org/documentation/refinement.htm#anch76 Note that you can't use mlhl target to refine f' and f'', so use ml instead. Pavel.