ImmuCellAI

ImmuCellAI estimates the abundance of immune cell types from gene expression data to quantify immune infiltration and analyze immunotherapy response in cancer.

Key Features:

• Gene set signature-based method: Uses curated gene set signatures to infer cell-type-specific abundance from gene expression profiles.
• Cell-type coverage: Estimates abundance of 24 immune cell types, including 18 distinct T-cell subsets.
• Input data: Operates on gene expression and sequencing data.
• Validation: Validated using sequencing data, flow cytometry results, and public expression datasets.
• Comparative accuracy: Demonstrated superior accuracy relative to existing methods in benchmarking described by the authors.
• Predictive modeling support: Enables development of immunotherapy response predictors with reported AUCs ranging from 0.80 to 0.91.

Scientific Applications:

• Tumor immune profiling: Quantifies tumor-infiltrating immune cells to characterize immune microenvironment composition.
• T-cell subset analysis: Profiles 18 T-cell subsets to assess their distribution and potential functional roles in cancer.
• Immunotherapy response analysis: Compares immune cell abundance between treatment stages (on-treatment vs pre-treatment) and between responders and non-responders, including dendritic cells (DC), cytotoxic T cells, and gamma delta T cells.
• Predictive biomarker development: Supports building models to predict immunotherapy outcomes with reported AUC 0.80–0.91.

Methodology:

Applies a gene set signature-based computational approach to gene expression profiles to estimate abundance of 24 immune cell types; validation used sequencing data, flow cytometry measurements, and public expression datasets.