The Microbe Directory: a centralized database for biological interpretation of microbiome data

Area of Science:

• Microbiology
• Bioinformatics
• Computational Biology

Background:

• Metagenomic data analysis requires comprehensive microbial metadata.
• Existing databases may lack detailed phenotypic and ecological traits.
• Manual curation is crucial for accurate biological interpretation.

Purpose of the Study:

• To introduce The Microbe Directory (TMD), a centralized database of microbial metadata.
• To provide an interactive browser for accessible data exploration.
• To compare manual curation with AI-driven annotation for microbiome data.

Main Methods:

• Centralized database creation with phenotypical and ecological traits.
• Manual annotation by trained student volunteers.
• Development of an interactive browser for data access.
• Analysis of microbiome samples from diverse projects (MetaSUB, TARA Oceans, HMP, Sponge Microbiome Project).

Main Results:

• TMD successfully integrates metadata for diverse microbial domains.
• The interactive browser enhances accessibility for non-bioinformaticians.
• Manual curation proved superior to AI for accurate microbial annotation, despite AI's speed.
• Demonstrated utility of TMD in analyzing real-world microbiome datasets.

Conclusions:

• TMD offers a unique, manually curated resource for interpreting microbiome data.
• Manual annotation by domain experts is essential for reliable biological insights.
• The database facilitates the discovery of biological associations within microbial communities.