Extended Data Fig. 3: Alignment of NPC scaffolds.
From: Overlapping nuclear import and export paths unveiled by two-colour MINFLUX
a, MINFLUX images of single NPCs. Images are 2D histograms of localizations for individual NPCs from permeabilized U2OS cells containing NUP96-mEGFP labeled with NbGFP-HMSiR. b, Scatter plot of HMSiR localizations. Only high-density circular clusters were analyzed further (squares); deformed or incomplete clusters were rejected. c, Double-circle fitting. High quality localization clusters were fit to a double-circle model7, reflecting the double-ring structure of the NPC. d, Rotation phase angle. The angles in the xy plane of the individual localizations in c relative to the centroid of the double-circle fit were binned (0–45°; assumes an eightfold periodicity), normalized, and fit to y = 1/9 + (1/20.6)*sin(8(x-ϕ)), as described previously7. The 1/9 term reflects the average frequency expected for the 9 bins (5° each), and the sine scaling factor is a reasonable average based on simulations. Note that improving the chi-square of the fit by allowing for an adjustable scaling factor does not change the estimated phase angle due to the orthogonality of frequency and angle. e, 2D histogram of aligned NPC scaffolds. Localization clusters were rotated by a phase angle, as determined in d, and aligned based on their centroids, as determined in c (37 cells, 541 NPCs, N = 82,331 localizations). This is the same image as in Fig. 1g. f, Angular distribution of rotationally corrected localizations. The angle distribution for the individual localizations in e was fit to y = 1/180 + csin(8(x-ϕ)), where c and ϕ are fit parameters, and the 1/180 term reflects the average frequency expected for the 180 bins (2° each). g, The number of HMSiR localizations obtained per NPC scaffold. As few as ten localizations were used earlier to identify an NPC scaffold7, but here more than 10-fold more localizations were obtained, on average.
