Comparative analysis of single-cell pathway scoring methods and a novel approach

Area of Science:

• Computational Biology
• Genomics
• Bioinformatics

Background:

• Single-cell gene set analysis (scGSA) is crucial for interpreting transcriptomic data.
• Existing scGSA methods face challenges including gene set size, quality, and data dropouts.
• Accurate pathway quantification is essential for understanding cellular functions.

Purpose of the Study:

• To introduce a novel method, single-cell Pathway Score (scPS), for robust gene set activity measurement at single-cell resolution.
• To benchmark scPS against six established scGSA tools.
• To identify critical factors influencing scGSA performance.

Main Methods:

• Development of the scPS algorithm for single-cell pathway scoring.
• Comparative analysis of scPS with AUCell, AddModuleScore, JASMINE, UCell, SCSE, and ssGSEA.
• Benchmarking using two simulation strategies evaluating cell count, gene set size, noise, and imputation effects.

Main Results:

• scPS demonstrates comparable performance to existing single-cell scoring methods.
• The scPS method identifies fewer false positives compared to other tested approaches.
• Simulation results highlight the impact of cell count, gene set size, and data noise on method performance.

Conclusions:

• scPS provides a reliable alternative for gene set activity analysis in single-cell studies.
• Understanding variables like gene set size and noise is critical for optimizing scGSA.
• This study contributes to improving the biological interpretation of complex single-cell transcriptomic data.