powmic: an R package for power assessment in microbiome case-control studies

Area of Science:

• Microbiology
• Bioinformatics
• Statistical Genetics

Background:

• Accurate sample size determination is crucial for identifying differentially abundant microbes in metagenomic sequencing.
• Traditional power analysis methods are often inadequate for complex microbial data characterized by excessive zeros, over-dispersion, and correlations.

Purpose of the Study:

• To develop a robust simulation-based power assessment strategy for metagenomic case-control studies.
• To address the challenges posed by complex microbial data characteristics in power analysis.

Main Methods:

• Developed a simulation-based power assessment strategy.
• Created the R package 'powmic' to implement the strategy.
• Considered key microbial data complexities: excessive zeros, over-dispersion, compositionality, intrinsic correlations, and variable sequencing depths.

Main Results:

• The 'powmic' R package provides a practical approach to power analysis for metagenomic data.
• The simulation-based strategy effectively accounts for the unique challenges of microbial data.

Conclusions:

• The 'powmic' package offers a valuable tool for researchers conducting metagenomic case-control studies.
• Accurate power analysis using 'powmic' can improve the reliability of identifying differentially abundant microbes and inform experimental design.