scParser: sparse representation learning for scalable single-cell RNA sequencing data analysis.

Area of Science:

• Computational biology
• Genomics
• Bioinformatics

Background:

• Single-cell RNA sequencing (scRNA-seq) data is rapidly increasing in volume and complexity.
• Existing integrative analysis methods often lack the scalability and focus needed to dissect heterogeneous biological condition effects across diverse cell populations.

Purpose of the Study:

• To develop a scalable computational approach, scParser, for the integrative analysis of scRNA-seq data.
• To model heterogeneous effects of biological conditions across cell subpopulations.
• To identify key gene expression mechanisms driving phenotypes and disease pathogenesis.

Main Methods:

• Developed scParser, a scalable method for integrative scRNA-seq data analysis.
• Implemented modeling of heterogeneous biological condition effects within cell populations.
• Extended scParser to pinpoint disease-contributing processes in specific cell subpopulations.

Main Results:

• scParser demonstrates favorable performance in cell clustering compared to existing state-of-the-art methods.
• The approach effectively models heterogeneous condition effects, revealing gene expression contributions to phenotypes.
• Extended scParser successfully identifies disease-related biological processes within cell subpopulations.

Conclusions:

• scParser provides a scalable and effective solution for integrative scRNA-seq analysis.
• The method enhances understanding of how biological conditions impact gene expression heterogeneously across cell types.
• scParser offers broad applicability for dissecting complex biological questions, including disease mechanisms.