propeller

propeller detects differences in cell type proportions across experimental conditions using single-cell RNA sequencing (scRNA-seq) data to identify statistically significant shifts in cellular composition.

Key Features:

• Statistical testing for cell type proportions: Detects statistically significant shifts in cell type composition between experimental groups using scRNA-seq data.
• Biological replication: Incorporates biological replication into the analysis framework to account for sample-level variability in scRNA-seq datasets.
• Variability modeling: Accounts for different sources of variability inherent in scRNA-seq data to improve robustness of proportion estimates.
• Validation by simulation: Performance and flexibility have been demonstrated through simulations with varied cell type proportion scenarios.
• Implementation: Implemented within the speckle R package.
• Support for complex designs: Designed to handle complex experimental designs that include multiple conditions or treatments.

Scientific Applications:

• Human Heart Development: Applied to investigate changes in cell type proportions during heart development.
• Aging Studies: Used to analyze how aging affects cell type composition.
• COVID-19 Disease Severity: Applied to scRNA-seq data from COVID-19 patients to identify shifts in cell type proportions associated with disease severity.

Methodology:

Leverages biological replication and explicit modeling of scRNA-seq variability to test for differences in cell type proportions across groups, with performance evaluated using simulations of varied cell type proportion datasets.