DCATS

DCATS performs differential composition analysis on single-cell omics data to identify cell types with statistically significant changes in abundance across experimental conditions.

Key Features:

• Beta-binomial regression framework: Uses a beta-binomial regression model to test for differences in cell-type composition between experimental states.
• Flexible experimental designs: Accommodates comparisons across multiple experimental conditions and complex study designs.
• Cell type assignment uncertainty modeling: Accounts for uncertainty in cell type labels during composition inference.
• Variance accommodation: Handles variability and uncertainty inherent to single-cell omics data.
• R package implementation: Provided as an implementation in R.
• Empirical performance: Demonstrated high sensitivity and specificity compared to state-of-the-art methods in evaluations.

Scientific Applications:

• Differential composition analysis: Identifying cell types with significant abundance changes across experimental conditions in single-cell omics studies.
• Multi-condition studies: Comparing cellular composition across multiple experimental states or complex designs.
• Analyses with uncertain annotations: Inferring composition when cell type assignments from clustering or annotation are uncertain.
• Cellular heterogeneity investigations: Quantifying shifts in cellular heterogeneity across biological contexts.

Methodology:

Applies a beta-binomial regression framework that models cell type assignment uncertainty and accommodates flexible experimental designs; implemented in R.