Cluster-independent multiscale marker identification in single-cell RNA-seq data using localized marker detector...

Area of Science:

• Genomics
• Computational Biology
• Bioinformatics

Background:

• Accurate cell marker identification is essential for understanding cellular diversity and function in single-cell RNA sequencing (scRNA-seq) data.
• Existing methods may struggle with fine-grained cellular distinctions and cross-sample comparisons.

Purpose of the Study:

• To introduce Localized Marker Detector (LMD), a novel computational tool for identifying "localized genes" in scRNA-seq data.
• To characterize cellular diversity at multiple resolutions and facilitate fine-grained analysis.
• To enable robust cross-sample comparisons without batch correction.

Main Methods:

• LMD constructs a cell-cell affinity graph to model cell similarity.
• Gene expression values are diffused across the cell graph.
• A diffusion dynamics-based scoring system identifies candidate marker genes.
• Candidate markers are grouped into functional gene modules.

Main Results:

• LMD successfully identified localized genes that accurately reflect cell types, subtypes, and cell cycle status.
• Application to mouse bone marrow and hair follicle data revealed shared and sample-specific gene signatures and novel cell populations.
• LMD demonstrated superior performance compared to eight existing methods across ten scRNA-seq datasets.

Conclusions:

• LMD is an effective tool for discovering cell markers and characterizing cellular heterogeneity in scRNA-seq data.
• The method facilitates robust cross-sample comparisons and identification of novel cell populations.
• LMD offers an improved approach for analyzing single-cell gene expression data.