The Use of Machine Learning and Explainable Artificial Intelligence in Gut Microbiome Research: A Scoping Review

Area of Science:

• Microbiome research
• Artificial intelligence in medicine
• Precision medicine

Background:

• Gut microbiome research increasingly utilizes machine learning (ML) and artificial intelligence (AI) for complex data analysis.
• Explainable AI (XAI) is vital for transparency and trust in clinical applications, particularly in precision medicine.
• Integrating XAI with gut microbiome data offers potential for more reliable clinical insights.

Purpose of the Study:

• To review ML and XAI applications in gut microbiome studies.
• To identify knowledge gaps and suggest future research directions.
• To evaluate the current state of explainability and reproducibility in the field.

Main Methods:

• Scoping review of publications from 2018-2024.
• Searched PubMed and Scopus databases.
• Selected 76 relevant publications for analysis.

Main Results:

• ML and AI are widely applied in gut microbiome studies, with a prevalence of 'black box' models like Random Forest.
• Explainability in the field is improving, with growing interest in techniques like SHAP (SHapley Additive exPlanations).
• Reproducibility of research remains an area needing greater attention.

Conclusions:

• Advancing XAI research in microbiome studies is essential for developing robust microbiome-based precision medicine.
• Increased focus on model transparency and reproducibility will enhance clinical translation.
• XAI holds significant promise for unlocking the full potential of gut microbiome data in healthcare.