ImmuCellAI-mouse

ImmuCellAI-mouse estimates the abundance of 36 immune cell (sub)types from mouse gene expression data to quantify immune-cell composition in mouse models.

Key Features:

• Hierarchical Prediction Strategy: Employs a three-layer hierarchical strategy to refine immune cell subtype abundance estimates.
• Robust Marker Gene Sets: Uses curated reference expression profiles and marker gene sets specific to mouse immune cell types.
• ssGSEA Enrichment Score Calculation: Predicts cell abundance by calculating single-sample Gene Set Enrichment Analysis (ssGSEA) enrichment scores from expression deviation profiles for each cell type.
• High Predictive Accuracy: Benchmarking shows correlation coefficients between predicted and actual cell proportions exceeding 0.8 for many cell types.
• Coverage of 36 Immune Cell (Sub)types: Provides quantitative estimates for 36 distinct immune cell (sub)types.
• Validation on Mouse Tumor Data: Applied and validated on a mouse breast tumor dataset, revealing dynamic changes in immune cell infiltration during treatment.

Scientific Applications:

• Immune Microenvironment Studies: Quantifies cellular composition and dynamics of the immune microenvironment in mouse models.
• Cancer Immunology Research: Characterizes immune cell roles in cancer progression and treatment response in mouse models.
• Immunotherapy Development: Supports evaluation and optimization of immunotherapeutic strategies in preclinical mouse studies.

Methodology:

Uses a three-layer hierarchical prediction strategy with curated reference expression profiles and marker gene sets, computing ssGSEA enrichment scores from expression deviation profiles for each cell type; performance was benchmarked by correlating predicted values with actual cell proportions.