[modeller_usage] bad occupancy from loop refinement

[albert <>]

Hello:
  I am using loop refinement protocol based on known PDB file method for
loop refinement. What surprised me is that: the occupancy for each atom
in the refined structure is more than 1. here is some list, can you tell
me why?

THX

-----------------------------Loop refinement----------------------------------------------
EXPDTA    THEORETICAL MODEL, MODELLER 9v8 2011/02/18 21:54:47
REMARK   6 MODELLER OBJECTIVE FUNCTION:       -11.8879
ATOM      1  N   SER A   1      53.405   4.206  73.937 59.86 59.86  
ATOM      2  CA  SER A   1      54.091   5.156  73.036 59.86 59.86  
ATOM      3  CB  SER A   1      53.358   5.255  71.687 59.86 59.86 
ATOM      4  OG  SER A   1      53.443   4.020  70.992 59.86 59.86
ATOM      5  C   SER A   1      54.148   6.519  73.639 59.86 59.86
ATOM      6  O   SER A   1      54.469   6.679  74.816 59.86 59.86
ATOM      7  N   ALA A   2      53.848   7.547  72.824 39.90 39.90
ATOM      8  CA  ALA A   2      53.862   8.892  73.313 39.90 39.90