PDF e-Pub

## Section: New Results

### Geo-colocalization and coorientation in fluorescence super-resolution microscopy

Participants : Frédéric Lavancier, Reda Alami Chantoufi, Aymeric Lechevranton, Antoine Salomon, Charles Kervrann.

Colocalization aims at characterizing spatial associations between two fluorescently-tagged biomolecules by quantifying the co-occurrence and correlation between the two channels acquired in fluorescence microscopy. This problem remains an open issue in diffraction-limited microscopy and raises new challenges with the emergence of super-resolution imaging. In [19], we proposed an original method (GcoPS) that exploits the random sets structure of the tagged molecules to provide an explicit testing procedure. GcoPS requires the adjustment of a $p$-value that guarantees more reproducibility and more objective interpretation and takes as inputs two 2D or 3D binary segmented images. This year, we extended this approach to the estimation of local co-localization. This amounts to applying the statistical test on windows randomly drawn in the whole image. A multiple testing procedure allows us to compute a global partial colocalization score. Meanwhile, the excursion sets of colocalization score map estimated by Gaussian smoothing are very helpful to detect regions of interest corresponding to significant colocalization and anti-colocalization sites. This approach has been evaluated on STORM (Stochastic Optical Reconstruction Microscopy) images which provides several hundreds thousands of super-localized positions of individual molecules with an average accuracy of 10-20 nanometers (see Figure 9). Finally, the method has been successively extended to the geo-coorientation (or geo-coalignment) of 2D-3D vectors (optical flow, tensors) and trajectories to analyze the molecular interactions.

Software: GcoPS (see Section 6.1).

Collaborators: J. Salamero (CNRS-UMR 144, Institut Curie, PSL Research University),

G. Bertolin (IGDR – Institute of Genetics & Development of Rennes),

M. Lelek and C. Zimmer (Institut Pasteur, Paris).

Figure 9. Illustration of geo-colocalization of two molecules (red/green channels) in STORM super-resolution microscopy (original image size: 4576 $×$ 3564 pixels; pixel size: 3 nanometers). (a) Overlay of two channels (sub-region of the original pair); (b) colocalization hits overlaid on the score (heat) map; (c) excursion sets of detected colocalization (white) and anti-colocalization (black) sites overlaid on the score map; (d) detected co-localization (circles) and anti-co-localization (squares) hits.