Hello,<br><br>We use iterative cycles of manual building (coot) - refinement (phenix) - density modification to improve maps for the next building step (RESOLVE).<br><br>Should we use FMODEL,PHIFMODEL from phenix or FCALC,PHIFCALC in the input to RESOLVE.  MODEL structure factors are closer to experimental ones but are they better for density modification purposes ? <br clear="all">

I think that if the program uses , as in DM, only Fobs and phases to calculate the maps then  PHIFCALC is better (no contribution of solvent) but if the programs calculates 2Fo-Fc maps then FMODEL,PHIFMODEL is better since it was shown that these maps are beter when bulk solvent correction is used.<br>
<br>Could you please comment whether this reasoning is correct and give information as to which is the case in resolve. <br><br>Thanks,<br>-- <br>Peter<br>