Re: [phenixbb] negative density for heavy atom?

Hi Kelly, I think Nigel's suggestion about the charge is very important and may explain why you have negative density (due to extra 6 electrons in a neutral state). So don't forget to add charge if it's not there yet. Refining occupancy may compensate for using incorrect charge, so I would ad charge first and if that doesn't help, then refine occupancy (or do it at once). Pavel. On 4/22/10 11:56 AM, Kelly Daughtry wrote:

The peak is directly over the WO4 molecule, overlaying pretty well with the 2Fo-Fc density.

In my refinement, I refined the B-factor. I will try to refine occupancy. That makes sense. If that atom is only there 50% of the time, there would be 50% less contribution, and thus I would see the negative density.

Thanks for the suggestions!

******************************************************* Kelly Daughtry PhD Candidate Department of Physiology and Biophysics Boston University School of Medicine 590 Commonwealth Ave R 390 Boston MA, 02215 (P) 617-358-5548 *******************************************************

On Thu, Apr 22, 2010 at 10:50 AM, Pavel Afonine mailto:[email protected]> wrote:

Hi Kelly,

- where is the peak: at atomic center, near it and looks like a bunch of concentric spheres? Then it might be a results of Fourier truncation ripples; check this out where I reply to a similar post:

http://proteincrystallography.org/ccp4bb/message14290.html

- did you try to refined f' and f'' of your heavy atom? Actually, I see you did (looking at .log file);

- did you try to refine its occupancy and anisotropic ADP?

If you send me (to my email address, not to the whole phenixbb) the data and model files, as well as the parameter file and exact command you used, I will have a look.

Pavel.

On 4/22/10 6:38 AM, Kelly Daughtry wrote:

...

Hello all, I have just used phenix autosol and autobuild to phase and build an initial model of my structure with phase information from Tungsten (W), collected at the peak wavelength of 1.2134 Angstroms. Autosol works beautifully! I was able to perform W-SAD and get great maps, when I initially thought I would have to perform W-MAD!

The problem I am having is during phenix.refine. When I run the refinement (with and without refining the group anomalous) and including my experimental phases, I see a giant negative peak in the Fo-Fc map at the location of my Tungsten. The rest of the map is beautiful. In the log file, I can tell that the scattering factor is high for tungsten (see below).

I believe the negative density is telling me I have too many electrons present at the tungsten site.

Can anyone help me figure out what I am doing wrong? I have attached my latest log file.

----------X-ray scattering dictionary----------

Number of scattering types: 7 Type Number sf(0) Gaussians W 1 73.70 2 Se 1 33.92 2 S 11 15.96 2 Mg 1 11.95 2 O 462 7.97 2 N 343 6.97 2 C 1326 5.97 2 sf(0) = scattering factor at diffraction angle 0.

========================== Anomalous scatterer groups =========================

Anomalous scatterer group: Selection: "name W" Number of selected scatterers: 1 f_prime: 16.87 f_double_prime: 17.72 refine: f_double_prime

Total number of atoms in anomalous groups: 1

******************************************************* Kelly Daughtry PhD Candidate Department of Physiology and Biophysics Boston University School of Medicine 590 Commonwealth Ave R 390 Boston MA, 02215 (P) 617-358-5548 ******************************************************* ------------------------------------------------------------------------ _______________________________________________ phenixbb mailing list [email protected] mailto:[email protected] http://phenix-online.org/mailman/listinfo/phenixbb

_______________________________________________ phenixbb mailing list [email protected] mailto:[email protected] http://phenix-online.org/mailman/listinfo/phenixbb

------------------------------------------------------------------------

_______________________________________________ phenixbb mailing list [email protected] http://phenix-online.org/mailman/listinfo/phenixbb