Huge thanks to @helucro.bsky.social and @markrobinsonca.bsky.social for great collaboration and supervision, and to the Robinsonlab and the @bioconductor.bsky.social community for valuable feedback throughout this project!

Feedback on the manuscript and package is welcome and much appreciated!

We highlight structure-based analysis using two publicly available datasets:
1. Quantifying structural rearrangements during colorectal malignancy transformation.
2. Recovery of structurally relevant gene expression gradients in human tonsil germinal centres.

Using our package sosta, we show how to reconstruct anatomical structures, quantify geometric features and other structurally-relevant characteristics, and compare features across samples and conditions.

Most spatial omics methods focus on single cells, but biological function often emerges from organised multicellular structures (like glands, crypts, or germinal centres)

I'm very excited to share our latest preprint!

We introduce structure-based analysis of spatial omics data – an approach that focuses on multi-cellular anatomical structures rather than single cells.

We also present sosta to facilitate this type of analysis: bioconductor.org/packages/sos...

🚀 Spatial Transcriptomics (Imaging) + Spatial-centric analyses by @helucro.bsky.social : the way to truly interrogate tissues!!!
📖 Read our pre-print on human tonsil architecture (full of tonsiLove ♥️) here: doi.org/10.1101/2025...

We are excited to share the publication of our paper on exploratory spatial statistics for spatial omics data

academic.oup.com/nar/article/...

Update: We greatly revised our paper and renamed it “Harnessing the Potential of Spatial Statistics for Spatial Omics Data with pasta”.

We discuss the broad range of exploratory spatial statistics options for spatial Omics technologies and show relevant use cases.

arxiv.org/abs/2412.01561

Many thanks to everyone involved 🤝 Martin Emons, @helucro.bsky.social , Izaskun Mallona, Reinhard Furrer, @markrobinsonca.bsky.social and all Robinsonlab members.

We give an overview of established methods for the analysis of both lattice- and point pattern-based data and discuss common challenges. More information can be found in the accompanying website: robinsonlabuzh.github.io/pasta/

In lattice-based analysis we assume that the locations were fixed at the time of sampling and study the associated features at each location accounting for spatial relationships. This offers an observation-based on the data.

Point pattern-based analysis offers an event-based view of the data. It allows us to study the processes that lead to an pattern that is e.g., clustered.

This offers two streams of analysis: point pattern- or lattice-based analysis.

Imaging-based data can be viewed as a point pattern either in terms of transcript locations or cell centroids. Alternatively, the segmented cell outlines can be interpreted as an irregular lattice. HTS-based approaches are most often recorded on a regular lattice.

Spatial omics data can be classified into imaging-based and high throughput sequencing (HTS)-methods that differ in resolution and the number of features targeted.

Interested in spatial statistics for spatial omics data? Check out our new resource: pasta.

We show how different technologies lead to different data modalities give and overview of point-pattern and lattice-based spatial analysis. Feedback welcome!

arxiv.org/abs/2412.01561