CircaPower

CircaPower calculates statistical power for detecting circadian rhythmicity in omics gene expression data by evaluating effects of sample size, intrinsic effect size, and sampling design.

Key Features:

• Statistical Methodology: Employs a theoretical framework for power calculation that explicitly accounts for sample size, intrinsic effect size, and sampling design.
• Simulation-Based Validation: Uses extensive simulations to quantify how variations in sample size, effect size, and sampling design affect statistical power.
• Cosinor Model Robustness: Fits data with a cosinor model and assesses robustness to violations of typical model assumptions.
• Real-World Data Examples: Demonstrated on mouse pan-tissue and human post-mortem brain datasets to illustrate performance on complex biological data.
• Case Study (RNA-Seq): Applied to mouse skeleton muscle RNA-Seq pilot data to exemplify experimental design and analysis considerations.

Scientific Applications:

• Experimental Design for Circadian Biology: Informs sample size and sampling design choices to ensure studies are adequately powered to detect circadian gene expression rhythms.
• Power Estimation for Omics Rhythmicity Detection: Provides quantitative power estimates for detecting rhythmicity in omics datasets, including RNA-Seq.
• Assessment of Model Sensitivity: Evaluates how cosinor-based detection sensitivity changes under varying effect sizes, sample sizes, and sampling designs.

Methodology:

Fits a cosinor model and performs simulation-based exploration of sample size, intrinsic effect size, and sampling design within a theoretical power-calculation framework.