MicrobiomeKG: bridging microbiome research and host health through knowledge graphs

Area of Science:

• Microbiome research
• Bioinformatics
• Computational biology

Background:

• The human microbiome plays a crucial role in host health.
• Understanding host-microbe interactions is key for health interventions.
• Existing research data is often heterogeneous and difficult to integrate.

Purpose of the Study:

• To develop MicrobiomeKG, a knowledge graph for microbiome research.
• To bridge microbial taxa and pathways with host health information.
• To facilitate hypothesis generation and advance personalized medicine.

Main Methods:

• Algorithmically extracted knowledge assertions from supplementary tables of microbiome papers.
• Neural networks for a standardized edge scoring system.
• Centrality analyses for identifying key relationships.

Main Results:

• MicrobiomeKG successfully links diverse microbial data with host health.
• Demonstrated use cases include helminth infections and NAFLD, Alistipes and inflammation, and Bifidobacterium and ADHD.
• Identified key microbial connections for various health conditions.

Conclusions:

• MicrobiomeKG provides a novel platform for integrative analysis in microbiome research.
• It facilitates the discovery of new biological relationships and aids hypothesis generation.
• The knowledge graph advances personalized medicine by elucidating microbial contributions to health and disease.