Figure 3
SHELXD substructure determination in thermolysin is more successful in data sets with higher (a) 〈I/σ(I)〉overall, (b) CC1/2 overall and (c) CCanom overall and (d) lower Rmeas overall. Because of the large number of data points (23 million), overlapping CCall/CCweak values are colour coded according to the density of points. Solutions are divided according to their merging values. The ranges are shown on the right-hand side of the plots. |