henhaobio
Sign in

bioRxivBioinformaticsPreprint

Voyager: exploratory single-cell genomics data analysis with geospatial statistics

Lambda MosesiD, Pétur Helgi EinarssoniD, Kayla C. JacksoniD, Laura LuebbertiD, A. Sina BooeshaghiiD, Sindri Emmanúel AntonssoniD, Nicolas BrayiD, Páll MelstediD, Lior S. PachteriD

PDFDOI
Structured summary available · Jump to summary ↓

Abstract

Exploratory spatial data analysis (ESDA) can be a powerful approach to understanding single-cell genomics datasets, but it is not yet part of standard data analysis workflows. In particular, geospatial analyses, which have been developed and refined for decades, have yet to be fully adapted and applied to spatial single-cell analysis. We introduce the Voyager platform, which systematically brings the geospatial ESDA tradition to (spatial) -omics, with local, bivariate, and multivariate spatial methods not yet commonly applied to spatial -omics, united by a uniform user interface. Using Voyager, we showcase biological insights that can be derived with its methods, such as biologically relevant negative spatial autocorrelation. Underlying Voyager is the SpatialFeatureExperiment data structure, which combines Simple Feature with SingleCellExperiment and AnnData to represent and operate on geometries bundled with gene expression data. Voyager has comprehensive tutorials demonstrating ESDA built on GitHub Actions to ensure reproducibility and scalability, using data from popular commercial technologies. Voyager is implemented in both R/Bioconductor and Python/PyPI, and features compatibility tests to ensure that both implementations return consistent results.

Abstract summary

Claude Opus 5.5 · xhigh effort · 18 Oct 2026 · from the abstract · CC BY-NC 4.0

Geospatial exploratory statistics, including local, bivariate and multivariate spatial methods, are packaged for spatial omics behind one interface, with matched R/Bioconductor and Python implementations.

Geospatial analyses refined over decades have not yet been fully adapted to spatial single-cell analysis, and the authors note that exploratory spatial data analysis is still not part of standard workflows. Voyager systematically brings that tradition to spatial omics, including local, bivariate and multivariate methods not yet commonly applied there, behind a uniform interface and a shared data structure.

Key results

  1. ClaimedVoyager exposes a set of spatial statistics, among them local, bivariate and multivariate methods the authors say are not yet common in spatial omics, through one consistent user interface.

    Described as the platform's contribution in the abstract; no list of individual statistics, benchmark or comparison against existing spatial omics toolkits is given. · Abstract

  2. ClaimedWorked examples are offered as evidence that these statistics surface biology, with biologically relevant negative spatial autocorrelation cited as a case in point.

    Asserted in the abstract without naming the tissue, genes or datasets involved, and without statistics supporting the biological relevance. · Abstract

  3. ClaimedThe platform is built on SpatialFeatureExperiment, a structure that holds geometries alongside gene expression, and ships as paired R and Python packages checked against each other.

    The structure is said to combine Simple Feature with SingleCellExperiment and AnnData; compatibility tests check that the two implementations agree, and tutorials run on GitHub Actions using data from commercial platforms. · Abstract

Methods, data and reproducibility

Voyager is a platform that carries exploratory spatial data analysis from geospatial statistics into spatial omics, offering local, bivariate and multivariate spatial methods through one interface. It rests on SpatialFeatureExperiment, which pairs Simple Feature geometries with SingleCellExperiment and AnnData, and ships as parallel R/Bioconductor and Python/PyPI packages with consistency tests between them.

  • datasets from widely used commercial spatial platforms, used in the tutorials
Methods & tools
Voyager, SpatialFeatureExperiment, Simple Feature, SingleCellExperiment, AnnData, Bioconductor, PyPI, GitHub Actions
Reproducibility gaps
Distribution is described (an R/Bioconductor package, a Python package on PyPI, tutorials built with GitHub Actions and tests comparing the two implementations), but no repository URL, package version or dataset accession is printed in the text.
Limitations (2)
  • Scalability and reproducibility are asserted through the tutorial build process, with no runtime, memory or dataset-size figures and no comparison against other spatial analysis packages. · noted in this summary
  • The biological payoff is illustrated by a single example type, negative spatial autocorrelation, with no tissue, gene or dataset named and no statistical support in the available text. · noted in this summary

Read this if you analyse spatial omics data and want exploratory geospatial statistics, including local and multivariate measures, available through one interface in both R and Python.

Citations
56

Reproducibility ledger

Nobody has reported trying to reproduce this paper yet. Did you replicate an experiment or run the code? Your report saves the next lab a week.

Sign in to file a reproducibility report.

Related papers

Similarity Map
SelectedSame fieldCross-field
THISMoses '25Pont '23Langseth '26Wang '23Yang '20Fernandez '26
73% semantic matchBioinformatics
Lambda Moses, Alik Huseynov, Joseph Matthew Rich+1 more·bioRxiv·Feb 24, 2025· 2 citations