[phenixbb] refinement of partly occupied water molecules?

Pavel Afonine pafonine at lbl.gov
Thu May 30 08:09:24 PDT 2013

Hi Stefan,

> in some of my macromolecular crystal structures with resolutions 
> between 1.1 - 1.4 Å, mainly in the second solvation shell round 
> positive Fo-Fc electron density blobs are detectable which show after 
> assignment of a water molecule to these blob and subsequent refinement 
> with Phenix refine a good-looking calculated 2Fo-Fc electron density 
> (round shape, the 2Fo-Fc map appears at a contour level of σ = 1.0 - 
> 1.4), but there also occurs a small negative Fo-Fc electron density 
> detectable inside the 2Fo-Fc density blob. The negative Fo-Fc electron 
> density disappears if the occupancy of the water molecule is refined 
> by Phenix or manually set lower than 100% occupancy. Therefore, I 
> think these positions are partly occupied by water molecules, but I am 
> not sure how I should handle it/how it is generally handled. I would 
> be thankful for any advice and/or literature about this topic.

currently phenix.refine can add automatically only fully occupied waters 
(I am not sure it will ever use atomic model for partially occupied 
waters). If necessary, add partially occupied waters yourself manually. 
Make sure they have properly set altlocs (alternative location 
identifiers): non blank (say A and B).

> My second question is concerning a density blob in perfect 
> coordination distance to an indolyl group (perfectly located in the 
> middle of the five-ring, with 3.4 Å distance to every of the five 
> carbon atoms of the five-ring). What‘s the best way to find out which 
> ion possibly occupies this position? I tried Na+ and Ca2+, but after 
> refinement big negative Fo-Fc blobs appear in both cases (maybe the 
> ion at this position is only partly occupied?). There is no anomalous 
> signal detectable at this position (data collection wavelength was 
> 0.91841 nm). After the refinement of a water molecule at this 
> position, a positive Fo-Fc election density is still there.

A manual way of doing this is to come up with a list of potential 
ion-candidates based on your knowledge of chemistry and what may be in 
crystallization solution, then try placing one by one, refine 
anisotropic ADP and occupancy. Make sure you do refinement until 
convergence, which means you may want to run about 10 or more 
macro-cycles. Finally, choose the one that results in flattest residual map.

I guess Nat's suggestion does something similar to this automatically.


More information about the phenixbb mailing list