Model-based and phylogenetically adjusted quantification of metabolic interaction between microbial species

Area of Science:

• Microbiology
• Computational Biology
• Systems Biology

Background:

• Microbial communities display complex interactions, including competition and cooperation.
• Inferring these interactions often uses co-occurrence data or genome-scale metabolic models.
• Phylogenetic similarity can bias interaction inference due to shared genomic and functional profiles.

Purpose of the Study:

• To develop a novel computational approach to estimate microbial competition and complementarity indices, adjusted for phylogenetic distance.
• To implement an automated pipeline, PhyloMint, for constructing these indices from metabolic models.
• To analyze bacterial interactions within the human gut microbiome, considering phylogenetic relationships.

Main Methods:

• Developed a novel approach to calculate competition and complementarity indices, accounting for phylogenetic distance.
• Implemented the PhyloMint pipeline using genome-scale metabolic models.
• Applied the pipeline to 2,815 human-gut associated bacterial genomes and performed network community analysis.

Main Results:

• A high correlation was observed between phylogenetic distance and metabolic competition/cooperation indices.
• Identified bacterial pairs with significantly higher cooperation scores than expected for their phylogenetic distance.
• Network analysis revealed distinct modules of bacterial interactions characterized by high cooperation and low competition.

Conclusions:

• Niche differentiation is a dominant factor shaping microbial interactions.
• Habitat filtering also contributes to interactions within specific bacterial clades.
• PhyloMint provides a robust method for analyzing phylogenetically informed microbial metabolic interactions.