MIDASim: a fast and simple simulator for realistic microbiome data.

Area of Science:

• Microbiome Research
• Bioinformatics
• Statistical Modeling

Background:

• Human microbiome sequencing reveals disease associations.
• Statistical methods are crucial for analyzing microbiome data.
• Simulating realistic microbiome data is essential for method validation.

Approach:

• Developed MIDASim (Microbiome Data Simulator), a two-step simulation approach.
• Generates correlated binary indicators for taxon presence-absence.
• Uses a Gaussian copula for relative abundances and counts, with parametric/nonparametric options.

Key Points:

• MIDASim accurately reproduces distributional and correlation structures of template microbiome datasets.
• Outperforms existing methods in capturing data features at presence-absence and relative-abundance levels.
• Handles complex distributional features with minimal assumptions.

Conclusions:

• MIDASim is a fast, simple, and flexible tool for microbiome data simulation.
• Its computational efficiency supports large dataset generation.
• Enhances validation and evaluation of microbiome statistical methods.