MOPower

MOPower simulates multi-omics datasets and performs statistical power calculations to support study design and detection of associations across the genome, transcriptome, epigenome, metabolome, proteome, and microbiome with clinical phenotypes.

Key Features:

• Data simulation: Simulates multi-omics data using statistical distributions across three different omic layers covering genome, transcriptome, epigenome, metabolome, proteome, and microbiome features.
• Power calculation: Calculates statistical power by analyzing multiple replicates using multi-omics analysis models and packages and supports large numbers of simulations across varying sample sizes.
• Integration methods: Supports testing of different integration models and methods, explicitly including MOFA (Multi-Omics Factor Analysis).
• Customization: Enables specification and customization of omics features and study design parameters for tailored simulation scenarios.
• Performance reporting: Reports computational performance metrics, including benchmarked runtimes (e.g., ~500 seconds for a representative power calculation run between integration models) that vary with feature counts and study design.

Scientific Applications:

• Rare disease study design: Enables planning and optimization of multi-omics studies with limited sample sizes typical of rare disease research by estimating power under different scenarios.
• Method evaluation: Allows comparison and assessment of multi-omics integration methods (for example MOFA) in terms of power to detect associations with clinical phenotypes.
• Study optimization: Facilitates selection of sample sizes and feature configurations to maximize the likelihood of detecting biologically meaningful associations across multiple omic layers.

Methodology:

Simulates multi-omics data using statistical distributions across three omic layers and computes power by analyzing multiple replicates with multi-omics analysis models and packages, including MOFA.