Figure 7
Experimental single antibody blinking data matches simulated data obtained for a 10 nm rod using SuReSim software. (A) A super-resolution GSD image of a cover slip coated with Alexa647-conjugated anti-rabbit secondary antibodies was imaged and the GSD image, unfiltered 3D point cloud and filtered point cloud, processed using filtering module 4 of the 3D SMLM Network Analysis pipeline (Fig. 1C) to retain blobs and remove noisy blinks, are shown. (B) Segmentation module 5 (Fig. 1C) was applied to the data to obtain the individual blobs and plots of #blinks per blob vs blob area are shown. Blobs with area ≤100 nm2 all had fewer than 6 blinks (red lines) and we removed all blobs with fewer than 6 blinks and with area ≤100 nm2. (C) Single and multiple antibody blobs were identified using K-means clustering (K = 2) based on the extracted blobs’ features. By identifying the single antibody blobs, we derive the blinking data from an isolated protein experimentally. (D) We used the derived blinking data from the real single antibody blobs to image a rod of 10 nm length using SuReSim software47. The table shows the average standard deviation for the spread of the blinks from the blobs for the experimental data derived from a single antibody and for the simulated (SuReSim) imaging of a 10 nm rod.
