journal menu
![[Figure 3]](ei5112fig3mag.jpg)
|
|
Figure 3
MatchMaps results for poorly isomorphous lysozyme data sets with a translation artifact. (a) Two views of the deposited coordinates for the ambient-pressure apo (4wld, red cartoon) and high-pressure GlcNAc-bound (4xen, gray cartoon, ligand as gray sticks) HEWL structures, which differ by a 1.48 Å translation (left to right from the view shown). (b) Two views following global alignment of the apo (red cartoon) and bound (blue cartoon, blue sticks) models, where the translation artifact disappears. Instead, the true structural change is revealed to be a slight constriction of the bound protein (up–down from the view shown). This constriction is 0.77 Å, as measured by the change in distance between Cα of residues 25 and 69. (c) An isomorphous difference map shows clear positive difference density (blue mesh, contoured at 2.5σ, carved at 1.75 Å from the ligand) around the ligand (gray sticks), but the density is weak and imprecise. (d) In contrast, MatchMaps at the same contour level (blue mesh, contoured at 2.5σ, carved at 1.75 Å from the ligand) shows high-resolution features of the ligand. (e) Carving the isomorphous map within 3 Å of the ligand reveals significant patches of both positive (blue) and negative (red) electron density with no clear structural correspondence. (f) The same visualization shows that MatchMaps is less noisy in this region. (g) Residues Cys64 and Cys80 are shown in red (apo, original coordinates) or dark gray (bound, original coordinates), with sulfur shown in yellow and other protein backbone in light gray. The positive (blue) and negative (red) isomorphous difference signals at 2.5σ correspond to the translation artifact between the unaligned coordinates. (h) Superposition of the bound model with the apo model (cysteines shown in blue). The MatchMaps density at 2.5σ is weaker and corresponds to the subtle vertical shift between the aligned structures. (i) and (j) Both structures contain a well-coordinated bound sodium ion. The sodium ion is shown as a purple sphere, the coordinating protein and water from the apo structure are shown as light-gray sticks, and hydrogen bonds are shown in yellow. (i) An Fo–Fc difference map computed using the bound structure factor amplitudes and the apo model, but with the sodium ion omitted. This map erroneously suggests that the apo and bound data differ at the position of the sodium ion, when in fact the signal derives from modeling error. (j) MatchMaps shows no signal for this modeling error, as desired. |
![[Figure 3]](ei5112fig3.jpg)
 | JOURNAL OF APPLIED CRYSTALLOGRAPHY |
ISSN: 1600-5767
access
