powmic

powmic estimates empirical statistical power for microbiome case-control studies by simulating metagenomic sequencing data to assess detectability of differentially abundant (DA) microbes.

Key Features:

• Simulation-based strategy: Employs a simulation-based approach to estimate empirical statistical power for microbiome case-control studies.
• Complex data characteristics: Incorporates excessive zeros, over-dispersion, compositional effects, intrinsic microbial correlations, and variable sequencing depths into simulations.
• Empirical power estimation: Simulates diverse scenarios to provide empirical estimates of statistical power for detecting differentially abundant microbes and informing sample size decisions.
• R package implementation: Implemented as an R package for conducting simulation-based power assessments on metagenomic sequencing data.

Scientific Applications:

• Differential abundance detection: Assess power to detect differentially abundant (DA) microbes between case and control groups in microbiome studies.
• Sample size determination: Inform sample size selection for metagenomic sequencing experiments to achieve desired statistical power.
• Evaluating data characteristics: Quantify the impact of excessive zeros, over-dispersion, compositional effects, microbial correlations, and variable sequencing depths on detectability of biological signals.

Methodology:

Uses simulation of metagenomic sequencing data that incorporate excessive zeros, over-dispersion, compositional effects, intrinsic microbial correlations, and variable sequencing depths to estimate empirical statistical power for detecting differentially abundant microbes.