CHETAH: a selective, hierarchical cell type identification method for single-cell RNA sequencing.

Area of Science:

• Genomics
• Computational Biology
• Bioinformatics

Background:

• Single-cell RNA sequencing (scRNA-seq) is revolutionizing life sciences.
• Accurate cell type identification is crucial for scRNA-seq data analysis.
• Existing methods may struggle with novel or ambiguous cell populations.

Purpose of the Study:

• To introduce CHETAH (CHaracterization of cEll Types Aided by Hierarchical classification), a novel algorithm for cell type identification.
• To evaluate CHETAH's accuracy, speed, and selectivity, particularly for challenging cell types.
• To demonstrate the utility of intermediate and unassigned cell categories.

Main Methods:

• CHETAH utilizes a hierarchical classification tree based on reference scRNA-seq data.
• It assigns cell types with a confidence score derived from gene expression variance.
• The algorithm incorporates options for intermediate or unassigned cell classifications.

Main Results:

• CHETAH achieves accuracy comparable to existing methods for known cell types.
• It demonstrates superior specificity in identifying unknown or malignant cells as intermediate/unassigned.
• Analysis of pancreas datasets revealed novel cell populations and continuum states.

Conclusions:

• CHETAH provides accurate and selective cell type identification for scRNA-seq data.
• The ability to assign intermediate/unassigned categories is vital for preventing misclassification.
• CHETAH offers valuable insights into unexplored tissues and complex cell populations.