IKAP_cells

IKAP_cells identifies major cell population groups and distinguishes differentially expressed (DE) genes in single-cell RNA-sequencing (scRNA-seq) datasets to support cell-type characterization.

Key Features:

• Systematic Parameter Tuning: IKAP_cells systematically tunes clustering parameters to optimize identification of major cell groups and to address the interdependency between clustering and differential expression.
• Identification of Major Cell Types: With default settings, IKAP_cells identifies key cell types such as T cells, B cells, natural killer cells, and monocytes across datasets including peripheral blood mononuclear cells and mouse cortex samples.
• Enhanced Differentiation of Cell Groups: The method identifies more distinguishing DE genes between groups compared to traditional approaches that vary clustering parameters.
• Recursive Application for Subtype Identification: IKAP_cells can be applied recursively within identified populations to detect subtypes and construct multi-layered cell ontologies.
• Automation and Efficiency: IKAP_cells automates the tuning of clustering parameters to accelerate scRNA-seq analysis workflows.
• Integration with Seurat: IKAP_cells integrates with the Seurat package and regresses out confounding variables such as mitochondrial gene counts and total UMI counts by default using Seurat's ScaleData, with alternative confounders configurable in Seurat metadata.

Scientific Applications:

• Immunology: Identification and characterization of major immune cell populations (T cells, B cells, natural killer cells, monocytes) from PBMC scRNA-seq datasets.
• Developmental Biology: Mapping cellular heterogeneity and constructing hierarchical cell ontologies via recursive subtype identification.
• Neuroscience: Delineation of major cell populations and subtypes in mouse cortex scRNA-seq datasets.

Methodology:

Systematic tuning of clustering parameters; identification of differentially expressed (DE) genes; recursive reclustering for subtype identification; regression of confounders (mitochondrial gene counts and total UMI counts) using Seurat's ScaleData; integration with Seurat.