[phenixbb] Two questions for a structure at 2.28 Å

Dear all, I am solving a structure at 2.28 Å with an ice ring around 3.3 Å. In this ice ring region, the completeness of data is reduced to 72%. Is this normal for the ice rings? Another question is that, when I do real space refinement the RMS deviation with Rwork/Rfree are increasing. Is it a good idea to ignore real space refinement if the all residues are fitting well in electron density? Your reply will be highly appreciated. Thanks a lot! With regards, Bishal Singh PhD student German cancer research center (DKFZ) University of Heidelberg, Germany