Fig. 3
From: A tessellation-based colocalization analysis approach for single-molecule localization microscopy
Colocalization analysis on well-characterized biological structures. a–f Colocalization analysis of 3D single-color Tubulin data. (a) Random split of the localizations to obtain 2 channels with a density ratio of 50/50 (scale bar 5 µm). Left: 2D projection overlay of the 2 channels. Right: 3D color coding projection of all the localizations. b Magnified views of 3 ROI in a (scale bar 2 µm). c Point rendering for different density ratios within the magnified view in b (left 33 : 67%, middle 20 : 80%, right 10 : 90%, scale bar 250 nm). d Scatterplot of the normalized densities of the 2 colors. e 2D colocalization analyses performed with Coloc-Tesseler (Manders and Spearman), Clus-DoC and Getis & Franklin. f 3D colocalization analysis performed with Coloc-Tesseler (Manders and Spearman). g–k Colocalization analysis of a 3D two-color Tubulin/Lamin dataset (scale bar 5 µm). h Lateral (top) and axial (bottom) views of two magnified ROI of the dataset in f (scale bar 2 µm). i Scatterplot of the normalized densities of the 2 colors. j 2D colocalization analyses performed with Coloc-Tesseler (Manders and Spearman), Clus-DoC and Getis & Franklin. k 3D colocalization analysis performed with Coloc-Tesseler (Manders and Spearman). l–o Colocalization analysis of 2D and 3D two-color acquisitions of nuclear pore complexes (NPC). l 3D NPC dataset of genome-edited expressing Nup107-SNAP cells, labelled with BG-AF647 (magenta) and WGA-CF680 (green) (scale bar 500 nm). m Magnified views of a 2D (left, middle) and 3D (right) NPC (scale bar 50 nm). The same 2D NPC was used for the intensity-based (left) and point (middle) rendering. The point representation (middle) illustrates the result of the 5-class classification obtained using Color-Tesseler. n Scatterplot of the normalized densities of the 2 colors for a 2D (left) and 3D (right) NPC data sets. o 2D (left) and 3D (right) colocalization analyses performed with Coloc-Tesseler (Manders and Spearman), Clus-DoC and Getis & Franklin. In all box plots the center line is the median, the square is the mean and the bounds of the boxes are the 75 and 25% percentiles i.e., the interquartile range (IQR)
