Partial Structure Factors

Suppose that one wants to freeze a major portion of the stucture during refinement. Under certain conditions, one can save CPU time by storing the calculated structure factors of the frozen part of the molecule and then adding them to the calculated structure factors of the free part of the molecule. The computation of $R$ values, crystallographic target functions, or electron density maps always refers to $F_c=F_{calc}+F_{part}$ (cf. Chapter 13). Note, however, that $F_{part}$ is not added to $F_{calc}$ when performing XREFin DO operations (Section 13.5).

In the following example, the structure factors of residues 1 and 2 are computed and written along with $F_{obs}$ to the file “partial.data":

 xrefin
   ...
   reflection @fobs.fob end
   selection=( resid 1:2 ) 
   update 
   do (FPART=FCALC)   
   write reflection fobs fpart output=partial.data end        
 end

In the next example, the partial structure factors are read, the structure factors of the remaining residues are computed, and $R$ for whole structure is computed:

 xrefin
   ...
   reflection reset @partial.data end
   selection=( not ( resid 1:2 ) ) 
   update
   print rfactor
 end

In addition, one should fix the frozen atoms during minimization and dynamics by including the statement

constraints fix=( resid 1:2 ) end

before running minimization or dynamics. One should use “method=direct" instead of “method=FFT" if the frozen part contains more than 80% of the molecule.