nf_xpatial: A Reproducible Framework for Standardized Preprocessing and Clustering of Xenium Data

Description:
Lara Ianov, Co-Director of the UAB Biological Data Science Core, University of Alabama at Birmingham

Spatial transcriptomics has rapidly emerged as a transformative technology in biomedical research, enabling the simultaneous characterization of gene expression and spatial organization of cells within intact tissues. Platforms such as 10X Genomics Xenium provide subcellular-resolution measurements that are critical for understanding tissue architecture, cellular heterogeneity, and microenvironmental interactions in health and disease. However, despite the growing adoption of Xenium datasets, standardized and reproducible computational workflows for their processing and exploration remain limited, often requiring ad hoc scripts and substantial bioinformatics effort.

Here, we present nf_xpatial, a modular and reproducible Nextflow based workflow designed for the comprehensive preprocessing workflow of Xenium spatial transcriptomics data. The workflow implements standardized analytical steps including comprehensive quality control (with support for cell area- or region-specific filtering), data normalization (area normalization and log normalization), multi-sample integration, and systematic parameter exploration for both single-cell clustering and spatially aware clustering approaches. By enabling iterative evaluation across multiple clustering resolutions and spatial modeling parameters, nf_xpatial facilitates robust identification of biologically meaningful spatial domains.

Overall, nf_xpatial streamlines the processing of Xenium data from raw inputs to integrated single-cell and spatial clustering outputs, offering a standardized starting point for biologists to evaluate, compare, and fine-tune parameters for downstream hypothesis-driven spatial analyses.

Nextflow Summit 2026 Boston
https://summit.nextflow.io/2026/boston/overview/