Characterizing tissue architecture at single-cell resolution - From (spatial) statistics to machine learning
Year: 2026
Type: dissertation
Abstract: Spatially structured cell heterogeneity within tissues is essential for healthy organ function. This heterogeneity arises from a tightly regulated interplay of cell proliferation, cell differentiation and spatial organization during development. In recent years, high throughput data acquisition techniques have been developed to study individual cell behavior within tissues at an unprecedented scale. These measurement techniques generate high-dimensional features per cells, known as morphometrics, transcriptomics and generally -omics data. These novel measures open the way for a renewed understanding of the relationship between the high-dimensional geometry of gene expression and the physical geometry of living tissue. I propose here a variety of approaches to assess the link between individual cell features and their global organization within tissue, both spatially by studying their organization into a spatial network, and temporally, by studying their dynamics within cell lineage trees. Our approaches cover a wide range of mathematical and computational tools, from (spatial) statistics to machine learning. The three main chapters discuss 1) single-cell and spatial morphometrics, 2) single-cell and spatial transcriptomics and 3) multi-modal data integration. Collectively, our results show that integrating findings from cell-scale feature patterning provides new insights into tissue development and organization. Finally, I propose new research avenues for the study of the high-dimensional geometry of mesoscopic structures in tissues, with a specific focus on the establishment and maintenance of boundaries.
Source: HAL
Author: Villoutreix, Paul
Language: English
Subfield: Molecular Biology
Domain: Life Sciences
Open Access status: green
OpenAlex ID: W7223190646