[phenixbb] Using LigandFit to identify unknown density

[Maia Cherney]

Hi Pavel,

I have six ligands at partial occupacies in my structure. Simultaneous 
refinement of occupancy and B factors in phenix gives a value of 0.7 for 
the ligand occupancy that looks reasonable.
How does phenix can perform such a refinement given the occupancies and 
B factors are highly correlated? Indeed, you can increase/decrease the 
ligand occupancies while simultaneously increacing/decreasing their B 
factors without changing the R factor value. What criteria does phenix 
use in such a refinement if R factor does not tell much?

Maia