Skip to main content
. 2017 Oct 23;6:e27131. doi: 10.7554/eLife.27131

Figure 2. Generation of 3D LocScale density maps and comparison of global and local sharpening procedures.

(A) Schematic illustration of the LocScale procedure for a Pol III density map. Two equivalent map segments (rolling windows) are extracted from the original 3D reconstruction of EMD-3180 (top) and a map simulated from the atomic model (PDB ID 5fja) (bottom). For each rolling window, the radially averaged structure factor profile is computed and used to scale the amplitudes of the corresponding window of the original reconstruction. Note, the fine structure of the model amplitudes has a characteristic profile in the resolution range <10 Å due to protein secondary structure and deviates from a simple exponential falloff (inset). (B) Effect of amplitude scaling illustrated for an exemplary density window. The density contained within a window cube is shown before (unsharpened) and after (LocScale) application of amplitude scaling. The central voxel of the rolling window is assigned the map value after amplitude scaling. The procedure is repeated by moving the window along the map until each voxel has been assigned a density value based on the locally estimated contrast. (C) Side views of TRPV1 densities obtained with different sharpening methods. Far left, EMD-5778 unsharpened; left, global Guinier B-factor −100 Å2; middle, local Guinier B-factor; right, unsharp masking and far right, LocScale sharpening. (D) Radial amplitude profiles of the respective maps shown in (C). (E) Mesh representation of the densities for transmembrane residues 430–449 superposed on the atomic model (color-coded by atomic B-factor). The order from left to right is the same as in (C) and (F). Side chains are shown in stick representation. (F) Mesh representation for a peripheral density region superposed on the atomic model shown in ribbon representation and color-coded by atomic B-factor. B-factor scale as in (E).

Figure 2.

Figure 2—figure supplement 1. Effect of secondary structure and atomic B-factors on the radial structure factor amplitude profile.

Figure 2—figure supplement 1.

(A) Typical amplitude profiles (magnified in (B)) computed from density maps simulated at 1 Å resolution of an idealized α-helix (pink), a representative β-sheet from PDB ID 3zqk (blue) and a random distribution of atoms (grey) (C). Note the characteristic differences (Debye effects) of the profiles of α-helical and β-sheet structures in comparison with the exponential decay from randomly distributed atoms. (D) Illustration on the amplitude profile dependence on the atomic B-factor for the density of an idealized α-helix (E). Note the gradual disappearance of the characteristic profile at high B-factors. Color-coding for profiles at different B-factors follows that in (E). (F) Side view (top) and cross-section of TRPV1 densities simulated from atomic models with a series isotropic B-factors. The disappearance of recognizable secondary structure periodicities in the real-space density correlates with the gradual disappearance of Debye effects in the Guinier plot and radial amplitude profiles in (G–H). (G) Guinier plot and (H) radially average amplitude profiles illustrating the B-factor dependence of Debye effect magnitude for the simulated TRPV1 channel maps. (I) Map kurtosis calculated for the simulated TRPV1 maps blurred with different overall B-factors as shown in (G–H).
Figure 2—figure supplement 2. Comparison of global and local sharpening methods for Pol III, γ-secretase and β-galactosidase.

Figure 2—figure supplement 2.

(A) Side view (top) and cross-section (bottom) of Pol III density maps obtained after sharpening with different sharpening methods as indicated. For comparison, detailed density features in the different maps are prominent for the central helix visible in the bottom part of the density map seen in the cross-section. In all cases (A–C) the left-most map represents the original 3D reconstruction before sharpening. Red arrows highlight density regions that are apparently different in the side view as an effect of the respective sharpening methods. (B) Comparison of sharpening methods applied to γ-secretase. (C) Comparison of sharpening methods applied to β-galactosidase. (D) Radial amplitude profiles for the maps shown in (A–C). The color coding in all cases follows that defined for Pol III (left).