CRAmed

CRAmed implements a conditional randomization test for high-dimensional mediation analysis in sparse microbiome data.

Key Features:

• Mediation effect decomposition: Decomposes the natural indirect effect into two parts corresponding to the presence-absence and abundance of a microbe, respectively.
• Performance evaluation: Demonstrated superior performance in Recall, precision, and F1 score compared to existing mediation analysis methods through comprehensive simulations.
• Robustness: Exhibits a notable level of robustness across simulation scenarios.
• Real data validation: Effectiveness and interpretability illustrated through two real data applications.

Scientific Applications:

• Microbiome mediation analysis: Infers the mediating role of the microbiome between treatment and outcome in human health and disease studies.
• Causal microbiome research: Supports exploration of the causal role of the microbiome in human complex traits and diseases.
• Host health factor identification: Uncovers factors influencing host health through microbiome-mediated pathways.

Methodology:

Conditional randomization test framework for high-dimensional mediation analysis.