Fig. 1: Applications of the Structural Repetition Detector (SReD) algorithm in fluorescence microscopy. | Nature Communications

Fig. 1: Applications of the Structural Repetition Detector (SReD) algorithm in fluorescence microscopy.

From: Structural Repetition Detector for multi-scale quantitative mapping of molecular complexes through microscopy

Fig. 1: Applications of the Structural Repetition Detector (SReD) algorithm in fluorescence microscopy.

a Detection of Structural Repetition Using Simulated Blocks: Microtubules imaged with STORM analysed for repetitive patterns using simulated structural blocks. The analysis was performed using SReD’s ‘block repetition’ mode, which features a ‘1-to-all’ matching scheme where a reference block is compared with all the remaining image blocks. A rotation-variant correlation metric (Pearson’s correlation coefficient) was used to account for structure orientation. Coloured regions in the repetition map correspond to repetitions of same-coloured blocks above. Scale bar: 2 µm. b Detection of Structural Repetition Using Empirical Blocks: HeLa cell nuclei stained with DAPI used to detect repetitive structural patterns using manually extracted empirical reference blocks. The analysis was performed in a similar manner as shown in (a) but the reference blocks are extracted directly from the input data. Coloured regions in the repetition map correspond to repetitions of same-coloured blocks in the previous subpanel. Scale bar: 30 µm. c Global Repetition Detection: Jurkat cell expressing inducible HIV Gag-EGFP fusion protein analysed using global repetition detection and a rotation-invariant metric (‘absolute difference of standard deviations’). The analysis was performed using SReD’s ‘global repetition’ mode, which features an ‘all-to-all’ matching scheme where all image blocks are compared with all the remaining image blocks. The repetition map reveals structures not easily detectable in input image and their relative frequency. Scale bar: 5 µm. d Multiscale Global Repetition: Xenopus laevis nuclear pores imaged with STORM analysed using different-sized receptive fields to detect structural repetition at various scales. The analysis was performed using SReD’s ‘global repetition’ mode, where each iteration used a different block-to-image size ratio. The repetition map identifies repeated structures from single nucleoporins (orange) to nucleoporin clusters (blue) and nuclear pore units (magenta). A rotation-invariant correlation metric was used to analyse structures irrespective of their orientation. Scale bar: 120 nm. Centre panel: Simplified SReD algorithm workflow, illustrating key steps from input preprocessing to repetition map generation.

Back to article page