scLAPA

scLAPA integrates gene expression and alternative polyadenylation (APA) information from single-cell RNA sequencing (scRNA-seq) to improve inference of cell-cell associations and cell-type identification despite high variability and dropout.

Key Features:

• Integration of Gene Expression and APA: Combines gene expression profiles and APA dynamics extracted from scRNA-seq to capture complementary transcriptomic information at single-cell resolution.
• Improved Cell-Cell Association Learning: Computes enhanced similarity metrics by integrating modalities, improving learning of cell-cell similarities and clustering accuracy, validated via comparative analyses using seven similarity metrics and five clustering methods across diverse scRNA-seq datasets.
• Discovery of Novel Cell Subpopulations: Enables detection of hidden subpopulations, exemplified by two peripheral blood mononuclear cell subpopulations not identifiable using gene expression alone.
• Cost-Efficient and Integrative Approach: Operates on existing scRNA-seq datasets without additional experimental modifications, leveraging APA information to augment downstream analyses.
• Comprehensive Toolkit: Provides strategies for learning cell-cell associations, improving cell-type clustering, and discovering novel cell types by integrating APA with gene expression.

Scientific Applications:

• Cell Type Clustering: Refines clustering of cell types by integrating APA data with gene expression for more accurate biological interpretation.
• Transcriptome Dynamics Analysis: Captures APA variations to study transcriptome dynamics across cell types.
• Novel Cell Population Discovery: Identifies previously overlooked cell subpopulations by combining APA and gene expression signals.

Methodology:

scLAPA processes scRNA-seq data to extract both gene expression and APA profiles and integrates these layers to compute enhanced similarity metrics for more robust clustering and association learning.