metamicrobiomeR

metamicrobiomeR analyzes microbiome relative abundance data and conducts study-level meta-analysis to test differential relative abundances and synthesize estimates across multiple studies.

Key Features:

• Zero-Inflated Beta GAMLSS: Applies Generalized Additive Models for Location, Scale, and Shape (GAMLSS) with a zero-inflated beta (BEZI) family to model relative abundance data that contain many zeros and are constrained to relative abundance scales.
• Differential Abundance Testing: Tests for differential relative abundances between comparison groups and provides estimates as log(odds ratios) of relative abundances.
• Meta-Analysis with Random Effects Models: Performs random effects meta-analysis to pool comparable estimates and their standard errors and to assess overall effects and heterogeneity across studies.
• Simulation Validation: Validates the performance of GAMLSS-BEZI through simulation studies.
• Real Data Application: Applies the methods to real microbiome datasets to demonstrate practical applicability.

Scientific Applications:

• Infant gut microbiome sex comparison: Compares gut microbiomes between male and female infants during the first six months of life.
• Cross-study synthesis: Integrates findings across multiple microbiome studies to derive consistent, pooled conclusions and evaluate heterogeneity.

Methodology:

Implements GAMLSS with a zero-inflated beta (BEZI) family to estimate log(odds ratios) for differential relative abundance testing; pools study-specific estimates and standard errors using random effects meta-analysis to assess overall effects and heterogeneity; validated by simulation studies and applied to real microbiome datasets.