Measuring the Similarity of SMLM-Derived Marked Point-Clouds

Session

Session 2 - Early Career Researchers Session: the Science of Tomorrow Today

Authors

Kylie Savoye (2), Mohammed Baragilly (2), Arthur Lewis (1), Fabian Spill (2), Dylan Owen (2)

Affiliations

1. AstraZeneca
2. University of Birmingham

Keywords

Single-Molecule Localisation Microscopy (SMLM)

Image Comparison and Image Analysis

Data Analysis 

3D Similarity Score

Clustering

Abstract text

Single-molecule localisation microscopy (SMLM) has transformed biological imaging by providing higher quality resolution at the nanoscale [1]. Unlike conventional microscopy, SMLM generates point-cloud data representing molecular coordinates rather than pixel arrays, offering insights into intricate biological structures [2]. Marked point-patterns, arising from SMLM combined with environmentally sensitive probes, associate each coordinate with biological properties, requiring more sophisticated analysis. To compare SMLM-derived point-clouds, existing methods using the Kolmogorov-Smirnov (KS) score neglect marked values (MVs), potentially overlooking significant biological interpretations associated with this 'extra' data. We propose an extended method integrating three KS scores: one focusing only on the data coordinates, another examining the MVs associated with the data, and a third utilising the sum of MVs. Comparing two point-clouds then yields a point in 3D space, where each KS score is a point coordinate, facilitating nuanced comparisons. Using a base-case, we evaluate our method against simulation data generated with varying parameters. Remarkably, point clouds generated with the same parameters as the base-case consistently exhibit the smallest distance to the origin, indicating high similarity. This approach promises deeper insights into complex biological data features. Understanding the spatial organisation of biomolecules is pivotal in elucidating cellular mechanisms. This method offers a robust framework for comparing SMLM-derived point-clouds, potentially unveiling novel biological insights.

References

[1] M. Lelek, M. T. Gyparaki, G. Beliu, F. Schueder, J. Griffie, S. Manley, R. Jungmann, M. Sauer, M. Lakadamyali, and C. Zimmer, “Single-molecule localization microscopy,” Nat. Rev. Methods Primers, vol. 1, June 2021.

[2] M. Baragilly, D. J. Nieves, D. J. Williamson, R. Peters, and D. M. Owen, “Measuring the similarity of SMLM-derived point-clouds.” Sept. 2022.