MetaTopics: an integration tool to analyze microbial community profile by topic model

Area of Science:

• Microbiology
• Bioinformatics
• Computational Biology

Background:

• Metagenomics studies face challenges in deciphering complex taxonomical structures from high-dimensional sequencing data.
• Existing workflows often fail to identify specific microbial communities and their association with disease status.
• Understanding the intricate relationships and interactions among bacteria within a microbial community remains largely unknown.

Purpose of the Study:

• To introduce MetaTopics, an interactive R package for analyzing and visualizing metagenomics taxonomy data.
• To efficiently extract latent microbial communities and quantify their influence on disease status.
• To provide a user-friendly tool for exploring microbe-microbe and microbe-disease associations.

Main Methods:

• Utilizes state-of-the-art topic models derived from statistical learning for data analysis.
• Employs the Quetelet Index for quantitative measurement of a sub-community's influence on disease status.
• Integrates analysis of microbial taxonomy with disease status for given samples.

Main Results:

• MetaTopics successfully extracts latent microbial communities, revealing intrinsic relations among major microbes.
• The Quetelet Index effectively quantifies the impact of microbial sub-communities on specific disease states.
• Demonstrated application on in-house oral and public gut metagenomics data highlights the package's utility.

Conclusions:

• MetaTopics is the first interactive R package to apply advanced topic modeling to metagenomics taxonomy data.
• The package facilitates the analysis and visualization of microbial communities and their disease associations.
• Provides a valuable resource for researchers studying the microbiome's role in health and disease.