Hi all I did phased small protein using SAD at 3.5 Ang data and able to build around 80 residues  as alanine and missing other half. So I do have a partial model and tried to used that to and combine with experimental phase and resulting map is not better than the experimental map alone. I do have other native data set slightly better resolution with different space group. I know my partial model is not enough to get any phase information in these data set.  Is it any possibility in phenix to improve the phase that I could help to model further? Any suggestions will be appreciated thanks venkat