[phenixbb] FW: Occupancy / alternate conformation Refinement

[Schulz, Eike-Christian]

    Dear all, 
    
    In one of my structures I appear to have a mixed state between a covalently bound intermediate and the free substrate occupying _almost_ the same position. I can model the states nicely in coot, but during refinement the molecules “fly” apart - although the occupancy was limited to 0.5 for their overlapping atoms.  
    
    I have tried to model the free ligand as an alternate conformation of the covalent intermediate: 
    
    HETATM  852  N   ASB A 110      10.523  15.702  10.022  1.00 14.48           N  
    HETATM  853  CA  ASB A 110       9.974  15.713   8.664  1.00 15.11           C  
    HETATM  854  C   ASB A 110       8.635  16.385   8.683  1.00 14.31           C  
    HETATM  855  O   ASB A 110       8.563  17.443   9.233  1.00 13.03           O  
    HETATM  856  CB  ASB A 110      10.053  14.307   8.091  1.00 11.72           C  
    HETATM  857  CG  ASB A 110       9.774  14.157   6.621  1.00 15.26           C  
    HETATM  858  OD1 ASB A 110      10.688  14.485   5.644  1.00 19.29           O  
    HETATM  859  OD2 ASB A 110       8.693  13.880   6.340  1.00 13.36           O  
    HETATM  860  C1 AASB A 110      10.637  15.649   3.353  0.50 27.61           C  
    HETATM  861  C2 AASB A 110      10.231  14.480   4.266  0.50 32.98           C  
    HETATM  862  O1 AASB A 110      11.505  16.454   3.690  0.50 31.63           O  
    HETATM  863  O2 AASB A 110      10.085  15.787   2.280  0.50 21.80           O  
    HETATM 4742  O2 BASB A 110      10.072  15.815   2.191  0.50 24.78           O  
    HETATM 4743  C1 BASB A 110      10.573  15.717   3.221  0.50 28.20           C  
    HETATM 4744  O1 BASB A 110      11.449  16.458   3.619  0.50 22.71           O  
    HETATM 4745  C2 BASB A 110      10.107  14.567   4.152  0.50 19.02           C  
    HETATM 4746  F  BASB A 110       9.287  13.735   3.610  0.50 34.13           F
    
    Although I have provided a new CIF file for the ligand named according to the alternate conformation this causes phenix to crash. 
    
    How can I correctly refine these overlapping atoms ?
    
    Thanks for your advice, 
    
    Eike