Immune cell identifier and classifier (ImmunIC) for single cell transcriptomic readouts.

Area of Science:

• Immunology
• Bioinformatics
• Computational Biology

Background:

• Single-cell RNA sequencing (scRNA-seq) is crucial for immune profiling, identifying disease markers, and therapeutic targets.
• Accurate immune cell-type annotation from scRNA-seq data is challenging due to gene expression heterogeneity.
• Existing computational methods for immune cell annotation lack high accuracy.

Purpose of the Study:

• To develop a simple and robust tool, ImmunIC, for accurate immune cell identification and classification.
• To improve the accuracy of immune cell annotation in complex multicellular samples.
• To advance cell-type specific immune profiling for disease signature and therapeutic target discovery.

Main Methods:

• ImmunIC combines marker genes with a machine learning approach for immune cell identification.
• The tool was trained and validated on over two million immune cells and half-million non-immune cells from 66 scRNA-seq studies.
• Performance was evaluated against existing immune cell classifiers.

Main Results:

• ImmunIC achieved 98% accuracy in identifying immune cells.
• The tool demonstrated 92% accuracy in categorizing cells into ten distinct immune cell types.
• Analysis of COVID-19 cases revealed peripheral blood mononuclear cell compositions and identified differential pathways.

Conclusions:

• ImmunIC is a highly accurate and robust tool for immune cell identification and classification.
• The tool enhances the utility of scRNA-seq data for immune profiling and discovery.
• ImmunIC can serve as a stand-alone bioinformatic cell sorter for advancing immune-related research.