POWSC

POWSC provides power evaluation and sample size estimation for single-cell RNA sequencing (scRNA-seq) experiments by estimating cell-type-specific parameters from real expression data and simulating zero-inflated scRNA-seq datasets to assess detection power for differentially expressed genes.

Key Features:

• Parameter Estimation: Estimates cell-type-specific parameters from real scRNA-seq expression datasets to parameterize simulations.
• Data Simulation: Simulates synthetic scRNA-seq data that incorporate zero-inflation and other characteristics observed in real datasets.
• Power Assessment: Performs stratified and marginal power analyses for differentially expressed (DE) genes, supporting phase-transition and magnitude-tuning definitions of DE.
• Sample Size Recommendation: Recommends sample sizes and optimizes trade-offs between sample number and sequencing depth while maintaining a constant total number of reads.

Scientific Applications:

• Experimental Design for scRNA-seq: Guides design of high-throughput scRNA-seq studies by providing power evaluations and sample size recommendations that account for sparsity and heterogeneity in single-cell data.

Methodology:

Estimates parameters from real expression datasets, simulates synthetic scRNA-seq data incorporating zero-inflation and real-data characteristics, conducts stratified and marginal power assessments for DE genes using phase-transition or magnitude-tuning, evaluates the relationships among power, sample size, and sequencing depth, and produces visualizations of these dynamics.