Label-free selection of marker genes in single-cell and spatial transcriptomics with geneCover

Area of Science:

• Genomics
• Bioinformatics
• Computational Biology

Background:

• Marker gene panel selection is crucial for analyzing single-cell RNA sequencing (scRNA-seq) and spatial transcriptomics data.
• Current label-based methods rely on predefined cell types, limiting their application.
• Existing label-free methods often fail to identify rare cell types or scale efficiently.

Purpose of the Study:

• Introduce geneCover, a novel label-free method for optimal marker gene panel selection.
• Address limitations of existing methods in scalability and identification of rare cell types.
• Leverage gene-gene correlations for robust marker gene identification.

Main Methods:

• Developed geneCover, a combinatorial, label-free approach.
• Utilized gene-gene correlations to select minimally redundant marker genes.
• Evaluated geneCover on diverse scRNA-seq and spatial transcriptomics datasets.

Main Results:

• geneCover demonstrates excellent scalability for large datasets.
• Identified marker gene panels effectively capture distinct transcriptomic correlation structures.
• Successfully distinguished cell states, including rare cell types, in various tissues.

Conclusions:

• geneCover offers a powerful, scalable, and label-free solution for marker gene selection.
• The method enhances the analysis of cellular and spatial heterogeneity in transcriptomic atlases.
• geneCover shows significant utility across diverse biological contexts and data types.