Skip to content

OSTA — Orchestrating Spatial Transcriptomics Analysis with Bioconductor

Open-source online book and companion data package — reproducible code examples and end-to-end workflows for spatial-omics analysis using Bioconductor in R, including interoperability with Python; the companion OSTA.data package ships example datasets across multiple spatial-omics platforms.

  • Maintainer (data package): Yixing E. Dong — estelladong729@gmail.com
  • Authors: Helena L. Crowell, Yixing Dong, Ilaria Billato, Peiying Cai, Martin Emons, Lukas Weber, Stephanie Hicks, Ellis Patrick, Mark Robinson, Vince Carey, others
  • Bioconductor (companion data): OSTA.data v1.4.0
  • Book: bioconductor.org/books/OSTA/
  • Source: github.com/lmweber/OSTA
  • License: Artistic-2.0
  • Status at EuroBioC2025: workshop / book in active development

Workshop at EuroBioC2025

  • Presenters: Yixing Dong & Ellis Patrick
  • Day / session: Day 2 (Thu Sep 18, 2025) — afternoon hands-on workshop (parallel with mia, msqrob2, Workshop Platform, iSEE)
  • Workshop title: "OSTA: Orchestrating Spatial Transcriptomics Analysis with Bioconductor"
  • Format: hands-on tutorial walking through book chapters using OSTA.data example datasets
  • Workshop materials: github.com/lmweber/OSTA and bioconductor.org/books/OSTA/
  • Video: workshops not recorded; the Bioconductor YouTube playlist covers plenaries only
  • Abstract / schedule: eurobioc2025 schedule

What it does

OSTA is an online textbook (in the OSCA / OHCA tradition) that walks through spatial-transcriptomics analysis end-to-end on the Bioconductor stack — preprocessing, QC, normalization, clustering, spatially-variable-gene detection, cell-type deconvolution, and cross-language interop with Python. The companion OSTA.data Bioconductor package provides the example datasets across multiple spatial-omics platforms (Visium, Xenium, MERFISH, etc.) so every chapter is runnable.

How it works

Core idea. OSTA is a Quarto-rendered book, not a package — its "mechanism" is editorial: it organizes the spatial-transcriptomics workflow into six parts (Background, Sequencing-based platforms, Imaging-based platforms, Platform-independent analyses, Multiple-sample analyses, Cross-platform analyses) so that the same dataset can be carried through preprocessing → QC → normalization → dimensionality reduction → clustering → spatial statistics → deconvolution → cell-cell communication chapters.

Inputs / outputs. Input for the reader: an R session with Bioconductor (the book reports being compiled on Bioconductor 3.23 / R 4.6.0). Input for the chapters: example datasets shipped by the OSTA.data companion package, loaded by name so every chapter is runnable without external downloads. Output: a SpatialExperiment (or platform-specific subclass) carried across chapters as the running object.

Key innovation. The book/data-package split is the operational design: chapters reference datasets by stable identifier, OSTA.data ships those datasets versioned on Bioconductor, and a Python-interop section bridges to the SpatialData / scverse stack so the book can teach R-side analysis without pretending the rest of the ecosystem doesn't exist. Splitting code (chapter) from data (package) keeps the book reproducible across Bioc releases.

Parameters worth knowing. Reader-facing decisions are mostly per-chapter (which clustering algorithm, which SVG method, which normalization). At the book level: which platform substrate to follow — sequencing-based (Visium / Visium HD) or imaging-based (Xenium, CosMx) — since the two halves diverge on QC and segmentation. MERFISH coverage not confirmed in the current table of contents (book lists Visium, Visium HD, Xenium, CosMx explicitly).

Canonical example. Each chapter pulls a worked dataset from OSTA.data; the sequencing-based section is anchored on Visium (including HD with binned and segmented variants), the imaging-based section on Xenium and CosMx. The multi-sample and cross-platform parts then re-use those same objects to demonstrate integration without re-importing raw data.

Where it fits in the corpus

  • AACR 2026: axis = single-cell & spatial omics — OSTA is the curricular spine for the R/Bioconductor side of the spatial-omics dossiers
  • DESpace (entry) — Peiying Cai is both the DESpace presenter and an OSTA author; OSTA is the natural place a DESpace tutorial would live
  • SpatialData (entry) — the Python-side framework that OSTA's interop chapters bridge to
  • Helena Crowell Day 1 keynote — overlaps directly; Crowell is both a keynote speaker and an OSTA author
  • ISMB 2026: scaffold; spatial-omics tracks
  • GBCC 2025: queued

Notes

OSTA is the consensus textbook around which much of the European spatial-omics Bioconductor work is converging. Tracking OSTA chapter additions is a good proxy for which spatial methods the Bioc community considers stable enough to teach.