Dear Pavel, Thank you for for your e-mail, Nigel (W. Moriarty) sent me exactly this link. I am copying my reply to him below, in case somebody else would be interested. "Thank you, actually the group occupancy refinement, together with refinement using anomalous data, seems to have solved the problem. The initial occupancy value for Tb was estimated from anomalous data, and the initial Ca occupancy at the same site then adjusted accordingly. Also, I had to set an almost zero-length "bond" between Ca and Tb to keep them at the same spot, it was a bit tricky to define correct distance (0.0 and sigma of 0.0 did not work, I then noticed that sigma of 0.0 is not OK)." Kind regards, Miha On 2025-03-07 00:58, Pavel Afonine wrote:

Hi Miha,

this fits into one of occupancy refinement scenarios described here:

https://phenix-online.org/phenixwebsite_static/mainsite/files/newsletter/CCN...

Let me know if you'd need any further help!

Pavel

On 3/4/25 06:44, [email protected] wrote:

...

Hi everyone, I have a dataset from a protein that contains 2 Ca2+ binding sites. The Ca2+ loaded protein was crystallized in the presence of Tb3+ which at one site mostly replaced Ca2+, however at the other site the Tb3+/Ca2+ ratio appears to be much lower (judging from anomalous map; data collected at 1.3 Å wavelength, MR for phasing). The occupancy sum of Tb3+ and Ca2+ at each of the two sites is strongly believed to be 1. While I am aware of the ion placement option in phenix.refine, I am not so sure how to refine the partial occupancies using anomalous data, if possible. Many thanks for any hints, Miha _______________________________________________ phenixbb mailing list -- [email protected] To unsubscribe send an email to [email protected] Unsubscribe: phenixbb-leave@%(host_name)s