[phenixbb] Fo-Fo map

Nathaniel Echols nechols at lbl.gov
Sat Jul 20 08:17:31 PDT 2013

On Fri, Jul 19, 2013 at 7:31 PM, Xiang Li <li338 at purdue.edu> wrote:
> I am working on two datasets of the similar crystal design. In one of the datasets, there is an extra guest protein (might be considered as a ligand) docked in one unit cell. The two crystals have very similar unit cell size and same symmetry. I want to calculate the Fo-Fo map to see whether I could figure out where the protein is. But I got error says the symmetry of the two datasets are not the same. I have never used the "isomorphous difference map" in GUI before. Can anyone help me to figure out where I did wrong please? I have attached the mtz file that I want to compare.

Your datasets differ by 3Å along the C edge, which is usually enough
to be considered "non-isomorphous".  At this low resolution, perhaps
it won't matter, but you should be very cautious about interpreting
the results, and pay special attention to the R-factors between the
datasets.  You can disable the symmetry check by clicking the box
labeled "Ignore non-isomorphous unit cells"; it will simply use the
symmetry from the first dataset.  (Which is what the phasing model
should also correspond to.)

> Btw, if the protein is not very rigidly linked with the crystal framework (~2 free residues linkage), am I supposed to see the density of the protein? Or are there any better ways for me to identify the protein?

It depends on how large and how well ordered it is, but the only way
to know for sure is to try.  If you don't see the density you're
expecting/hoping for, there could be multiple explanations.  As
always, having an anomalous scatterer helps a lot (but at this
resolution, sulfur certainly doesn't count).


More information about the phenixbb mailing list