Learning tissue representation by identification of persistent local patterns in spatial omics data

Abstract Spatial omics data provide rich molecular and structural information on tissues. Their analysis provides insights into local heterogeneity of tissues and holds promise to improve patient stratification by associating clinical observations with refined tissue representations. We introduce Ka...

Full description

Saved in:
Bibliographic Details
Main Authors: Jovan Tanevski, Loan Vulliard, Miguel A. Ibarra-Arellano, Denis Schapiro, Felix J. Hartmann, Julio Saez-Rodriguez
Format: Article
Language:English
Published: Nature Portfolio 2025-04-01
Series:Nature Communications
Online Access:https://doi.org/10.1038/s41467-025-59448-0
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850206484910047232
author Jovan Tanevski
Loan Vulliard
Miguel A. Ibarra-Arellano
Denis Schapiro
Felix J. Hartmann
Julio Saez-Rodriguez
author_facet Jovan Tanevski
Loan Vulliard
Miguel A. Ibarra-Arellano
Denis Schapiro
Felix J. Hartmann
Julio Saez-Rodriguez
author_sort Jovan Tanevski
collection DOAJ
description Abstract Spatial omics data provide rich molecular and structural information on tissues. Their analysis provides insights into local heterogeneity of tissues and holds promise to improve patient stratification by associating clinical observations with refined tissue representations. We introduce Kasumi, a method for identifying spatially localized neighborhood patterns of intra- and intercellular relationships that are persistent across samples and conditions. The tissue representation based on these patterns can facilitate translational tasks, as we show for stratification of cancer patients for disease progression and response to treatment using data from different experimental platforms. On these tasks, Kasumi outperforms related approaches and offers explanations of spatial coordination and relationships at the cell-type or marker level. We show that persistent patterns comprise regions of different sizes, and that non-abundant, localized relationships in the tissue are strongly associated with unfavorable outcomes.
format Article
id doaj-art-9c847f57e9cc4e55b6e754053b11e50d
institution OA Journals
issn 2041-1723
language English
publishDate 2025-04-01
publisher Nature Portfolio
record_format Article
series Nature Communications
spelling doaj-art-9c847f57e9cc4e55b6e754053b11e50d2025-08-20T02:10:49ZengNature PortfolioNature Communications2041-17232025-04-0116111510.1038/s41467-025-59448-0Learning tissue representation by identification of persistent local patterns in spatial omics dataJovan Tanevski0Loan Vulliard1Miguel A. Ibarra-Arellano2Denis Schapiro3Felix J. Hartmann4Julio Saez-Rodriguez5Institute for Computational Biomedicine, Heidelberg University and Heidelberg University HospitalInstitute for Computational Biomedicine, Heidelberg University and Heidelberg University HospitalInstitute for Computational Biomedicine, Heidelberg University and Heidelberg University HospitalInstitute for Computational Biomedicine, Heidelberg University and Heidelberg University HospitalSystems Immunology and Single-Cell Biology, German Cancer Research Center (DKFZ)Institute for Computational Biomedicine, Heidelberg University and Heidelberg University HospitalAbstract Spatial omics data provide rich molecular and structural information on tissues. Their analysis provides insights into local heterogeneity of tissues and holds promise to improve patient stratification by associating clinical observations with refined tissue representations. We introduce Kasumi, a method for identifying spatially localized neighborhood patterns of intra- and intercellular relationships that are persistent across samples and conditions. The tissue representation based on these patterns can facilitate translational tasks, as we show for stratification of cancer patients for disease progression and response to treatment using data from different experimental platforms. On these tasks, Kasumi outperforms related approaches and offers explanations of spatial coordination and relationships at the cell-type or marker level. We show that persistent patterns comprise regions of different sizes, and that non-abundant, localized relationships in the tissue are strongly associated with unfavorable outcomes.https://doi.org/10.1038/s41467-025-59448-0
spellingShingle Jovan Tanevski
Loan Vulliard
Miguel A. Ibarra-Arellano
Denis Schapiro
Felix J. Hartmann
Julio Saez-Rodriguez
Learning tissue representation by identification of persistent local patterns in spatial omics data
Nature Communications
title Learning tissue representation by identification of persistent local patterns in spatial omics data
title_full Learning tissue representation by identification of persistent local patterns in spatial omics data
title_fullStr Learning tissue representation by identification of persistent local patterns in spatial omics data
title_full_unstemmed Learning tissue representation by identification of persistent local patterns in spatial omics data
title_short Learning tissue representation by identification of persistent local patterns in spatial omics data
title_sort learning tissue representation by identification of persistent local patterns in spatial omics data
url https://doi.org/10.1038/s41467-025-59448-0
work_keys_str_mv AT jovantanevski learningtissuerepresentationbyidentificationofpersistentlocalpatternsinspatialomicsdata
AT loanvulliard learningtissuerepresentationbyidentificationofpersistentlocalpatternsinspatialomicsdata
AT miguelaibarraarellano learningtissuerepresentationbyidentificationofpersistentlocalpatternsinspatialomicsdata
AT denisschapiro learningtissuerepresentationbyidentificationofpersistentlocalpatternsinspatialomicsdata
AT felixjhartmann learningtissuerepresentationbyidentificationofpersistentlocalpatternsinspatialomicsdata
AT juliosaezrodriguez learningtissuerepresentationbyidentificationofpersistentlocalpatternsinspatialomicsdata