Hi Ezequiel,

I added a Cu(II) ion in a blob of density using Coot (this metal is used in crystallization solution). When I refined using phenix, I noted that occupancy does not change from the arbitrary value that I set in Coot,

if this value greater than zero and less then one, and you did enable occupancy refinement, then the occupancy should be refined. Is that what you did and still get occupancy unrefined?

and negative density remains when refinement is complete.

Not surprisingly if occupancy was not refined. So please make sure you do refine occupancy.

When I manually adjust occupancy to 0.3,

No need to do so as phenix.refine will normally refine it.

phenix changes it to 0.29,

Ok, well, this contradicts to what you wrote above: occupancy does change during refinement.

and no negative density is seen, however B-factor does not move from my arbitrary setting of 30 in Coot. I worry if something is wrong in my refinement strategy. I thank any suggestion.

Do you refine it?

Some additional information: my mtz is merged F's, resolution is 1.3 A, refinement strategy includes XYZ coordinates, Occupancies and Individual B-factors.

No need to ask this as it is obvious, but just on case: you do refine anisotropic ADPs at this resolution, correct?

This may be worthwhile checking:

- Occupancy and ADP refinement sections of phenix.refine documentation:
https://www.phenix-online.org/documentation/reference/refinement.html

- Pages 8-14 here:
https://www.phenix-online.org/presentations/faq.pdf

If you are still having troubles or questions please let me know. I may need input data and model files to investigate further.

Pavel