Construction and analysis of co-occurrence network in the gut microbiome

Abstract:

[Objective] To explore the networked and the topological structure of gut microbiota, we applied network analysis in this study to characterize the gut microbiome co-occurrence networks. [Methods] Gut microbiome data were divided into two groups based on the equol-metaboliting ability of hosts. We constructed the co-occurrence network of gut microbiota with Spearman correlation coefficients with FDR judgment in each group and analyzed the difference between groups. At the same time, the topological structure of random network was used to compare with the real network to uncover the significant differences. Finally, the species taxonomy information was taken into the network and revealed different features.[Results] The networks of two groups retained 45 and 47 different species respectively and show different complexity. From our data, we found the structure of the real network topology is specific and more interaction within different phylum in equol producer group. [Conclusion] By network analysis, we can discover the complexity of the interactions among the different species of gut microbes, and demonstrate the feature of network topology that was rarely reported before. And the method will also provide a new perspective of gut microbiota research in the future.

Key words:gut microbiota;network analysis;stochastic network;equol