AMDA

AMDA performs adaptive multivariate two-sample testing to detect differential abundance of microbial taxa between two sample groups in microbiome studies.

Key Features:

• Multivariate Approach: Evaluates overall shifts in microbial community composition by jointly testing multiple taxa to capture interactions and correlations among taxa.
• Adaptive Testing Framework: Employs an adaptive testing framework to increase statistical power while controlling type I error rates in high-dimensional microbiome data.
• Simulation Studies and Real Data Validation: Has been evaluated using simulation studies and applied to real datasets, demonstrating improved power relative to existing methods.
• Control for Multiple Testing: Accounts for complex dependencies among taxa to mitigate issues arising from multiple comparisons and stringent correction procedures such as false discovery rate control.

Scientific Applications:

• Disease Association Studies: Identifies shifts in microbiota composition linked to diseases such as inflammatory bowel disease, obesity, and cancer.
• Environmental Microbiome Research: Analyzes changes in microbial communities in response to environmental factors or interventions.
• Personalized Medicine: Investigates associations between individual microbiome profiles and treatment responses or health outcomes.

Methodology:

Compares taxa composition between two conditions using an adaptive multivariate statistical test that accounts for high dimensionality and complex correlations among taxa while controlling type I error under multiple testing (e.g., false discovery rate).