Abstract
Microbiome studies estimate the functions of bacterial flora in situ on the basis of species composition and gene function; however, estimation of interspecies interaction networks is challenging. This study aimed to develop a method to predict the interaction networks among bacterial species from human gut metagenome data using bioinformatics methods. Our proposed method revealed that adjacent gene pairs involved in bacterial interspecies interactions are localized at boundary regions and encode membrane proteins mediating interactions between the intracellular and extracellular environments, e.g., transporters and channel proteins, and those mediating interactions between metabolic pathways. Actual human gut metagenome data displayed numerous such highly reliable interspecies interaction gene pairs in comparison with random simulated metagenome data sets, suggesting that the species composition of the actual microbiome facilitated more robust interspecific interactions. The present results indicate that molecular interaction networks in human gut flora are organized by a combination of interaction networks common to all individuals and group-specific interaction networks.
Highlights
Over a decade has passed since the culmination of large-scale microbiome studies [1,2,3,4,5,6,7]
Correlations among co-occurrence profiles were determined, because metabolism as an interspecific interaction can be realized only if these gene pairs are present in the human gut of the same individual
This study evaluated interspecific interaction networks of the human gut microbiome based on the scoring method developed with bacterial interaction pairs, displaying robust bacterial species interactions in the actual human gut environment
Summary
Over a decade has passed since the culmination of large-scale microbiome studies [1,2,3,4,5,6,7]. Molecular interactions may mediate common functions in the human microbiome even for combinations of different species. Databases of biological pathways and networks are typically used to obtain information regarding such molecular interactions among bacteria. KEGG is one of most popular databases, presenting large-scale data regarding molecular interactions [11], and is widely used as a promising pathway database to observe and predict bacterial interactions. The STRING database contains data regarding molecular interactions from various sources [12], providing information regarding various intermolecular interaction networks including
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
