Related Experiment Video
Updated: May 2, 2026
10:22
Comprehensive Spatial Profiling of Species-agnostic Transcriptomes via Stereo-seq
Published on: October 31, 2025
921
Resolving sensitivity, specificity and signal contamination in Xenium spatial transcriptomics
Mariia Bilous1, Daria Buszta1,2, Jonathan Bac1
1Biomedical Data Science Center, Lausanne University Hospital, University of Lausanne, Lausanne, Switzerland.
Nature Methods
|April 30, 2026
Summary
This study benchmarks Xenium spatial transcriptomics data quality, revealing technical noise. A new method, SPLIT, refines Xenium data by correcting transcript contamination, improving cell-type resolution and uncovering hidden T-cell exhaustion signatures.
Area of Science:
- Molecular Biology
- Genomics
- Biotechnology
Background:
- Spatial transcriptomics offers high-resolution gene expression mapping in tissues.
- Xenium is a popular platform, but its data properties and limitations require thorough characterization.
- Existing datasets lack comprehensive analysis of Xenium-derived data quality and technical noise.
Purpose of the Study:
- To comprehensively evaluate Xenium spatial transcriptomics data quality and limitations.
- To identify and quantify sources of technical noise, such as transcript spillover.
- To introduce a novel method for refining spatial transcriptomic data purity and resolution.
Main Methods:
- Profiling over 40 breast and lung tumor sections using Xenium spatial transcriptomics with diverse gene panels.
- Systematic dissection of technical noise, including transcript spillover, assay specificity, and panel performance.
- Utilizing single-nucleus RNA sequencing for precise quantification of transcript contamination.
- Development and application of the SPLIT (Spatial Purification of Layered Intracellular Transcripts) method.
Main Results:
- Characterization of one of the most comprehensive Xenium datasets to date.
- Demonstration of transcript contamination quantification using single-nucleus RNA sequencing.
- SPLIT method successfully improved signal purity, background correction, and cell-type resolution.
- Revelation of T-cell exhaustion signatures linked to malignant cell colocalization, previously obscured by noise.
Conclusions:
- Established a critical benchmark for Xenium spatial transcriptomics performance.
- Introduced SPLIT as a scalable strategy for refining spatial transcriptomic data.
- Highlighted the importance of addressing technical noise for accurate biological interpretation in spatial transcriptomics.
