Spotsweeper-py: spatially-aware quality control metrics for spatial omics data in the Python ecosystem

Area of Science:

• Computational Biology
• Bioinformatics
• Genomics

Background:

• Spatially-resolved transcriptomics (SRT) generates complex datasets.
• Global quality control (QC) metrics can inaccurately remove biological signals or miss localized artifacts in SRT data.
• Existing spatially-aware QC tools are limited to the R programming language, hindering integration with the Python/scverse ecosystem.

Purpose of the Study:

• To introduce SpotSweeper-py, a Python package providing neighborhood-aware QC metrics for SRT data.
• To enable seamless integration of advanced local QC within the Python/scverse environment.
• To improve the accuracy and reliability of SRT data analysis by reducing false positives and preserving tissue architecture.

Main Methods:

• Development of SpotSweeper-py, a Python package implementing neighborhood-aware z-scores for QC metrics.
• Computation of z-scores for total counts, log total counts, detected genes, and mitochondrial percentage.
• Demonstration of SpotSweeper-py performance on 10x Genomics Visium and VisiumHD datasets.
• Inclusion of plotting utilities for outlier visualization.

Main Results:

• SpotSweeper-py effectively computes local, spatially-aware QC metrics.
• The package integrates smoothly with the Python/scverse ecosystem.
• It successfully reduces false positives from global QC while preserving tissue-specific structures.
• Performance was validated on public Visium and VisiumHD datasets.

Conclusions:

• SpotSweeper-py provides a robust, Python-based solution for local QC in SRT analysis.
• This tool enhances the reliability of SRT data processing pipelines.
• It makes advanced, spatially-aware QC accessible to a wider range of researchers.