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