Uncovering disease-related multicellular pathway modules on large-scale single-cell transcriptomes with scPAFA

Area of Science:

• Computational biology
• Genomics
• Bioinformatics

Background:

• Pathway analysis of single-cell RNA sequencing (scRNA-seq) data is vital for disease research, providing biological insights.
• Existing tools for cell-level pathway activity scores (PAS) are computationally inefficient for large datasets.
• Current methods often overlook cross-cell type pathway patterns crucial for understanding complex diseases.

Purpose of the Study:

• To introduce single-cell pathway activity factor analysis (scPAFA), a Python library for efficient PAS computation on large-scale scRNA-seq data.
• To enable the discovery of biologically interpretable, multicellular pathway modules reflecting disease-related PAS alterations across cell types.
• To address the computational limitations and single-cell type focus of current pathway analysis tools.

Main Methods:

• Development of scPAFA, a Python library for rapid PAS computation.
• Implementation of factor analysis to identify multicellular pathway modules.
• Application to large-scale colorectal cancer (CRC) and lupus scRNA-seq datasets (over 1.2 million cells).

Main Results:

• scPAFA achieved over 40-fold reduction in PAS computation runtime compared to existing methods.
• Identified reliable and interpretable multicellular pathway modules in CRC and lupus datasets.
• These modules effectively captured disease heterogeneity and transcriptional abnormalities.

Conclusions:

• scPAFA significantly enhances computational efficiency for pathway analysis in large scRNA-seq datasets.
• The library facilitates the discovery of multicellular pathway modules, offering deeper biological insights into diseases.
• scPAFA is a valuable tool for advancing disease research, biomarker discovery, and understanding complex biological mechanisms at the pathway level.