journal menu
![[Figure 1]](ei5112fig1mag.jpg)
|
|
Figure 1
A demonstration of how structure factors depend sensitively on isomorphism. The E. coli DHFR data set 1rx2 is compared with a highly isomorphous structure (1rx1) and with a poorly isomorphous structure (1rx4, see also Fig. 2). (a) Correlation coefficients for the isomorphous pair (triangles) and poorly isomorphous pair (circles). Correlations were computed between structure factor amplitudes (solid lines) and cosines of structure factor phases (dashed lines), aggregated per resolution bin. A figure of merit (mean of cosine of differences) was computed between corresponding phases (dotted lines). While the isomorphous data correlate well even at high resolution, the poorly isomorphous data are uncorrelated even at moderate resolution. (b) and (c) Structure factor phases appear (b) highly correlated for the isomorphous structures but (c) mostly uncorrelated between poorly isomorphous structures. (d) and (e) For an isomorphous difference map to be meaningful, structure factor amplitudes should only differ when the phase difference is small, and structure factor phases should only differ when the amplitude difference is small (Rould & Carter, 2003  ). This requirement is met (d) in the isomorphous case but not (e) in the non-isomorphous case. In all panels, computed phases are obtained from the `PHIC' column of the deposited MTZ files. |
![[Figure 1]](ei5112fig1.jpg)
 | JOURNAL OF APPLIED CRYSTALLOGRAPHY |
ISSN: 1600-5767
access
