Strong associations between microbe phenotypes and their network architecture

TL;DR

This study investigates how the structure of microbial metabolic networks correlates with their phenotypes, using topological metrics and hierarchical modeling to identify significant associations across 32 species.

Contribution

It introduces an automated method to analyze network topology and its relation to microbial phenotypes, highlighting key metrics and subgroups relevant to biological functions.

Findings

Topological metrics strongly associate with microbial phenotypes.

Hierarchical linear modeling identifies relevant metric subgroups.

Method applicable to other network-based disciplines.

Abstract

Understanding the dependence and interplay between architecture and function in biological networks has great relevance to disease progression, biological fabrication and biological systems in general. We propose methods to assess the association of various microbe characteristics and phenotypes with the topology of their networks. We adopt an automated approach to characterize metabolic networks of 32 microbial species using 11 topological metrics from complex networks. Clustering allows us to extract the indispensable, independent and informative metrics. Using hierarchical linear modeling, we identify relevant subgroups of these metrics and establish that they associate with microbial phenotypes surprisingly well. This work can serve as a stepping stone to cataloging biologically relevant topological properties of networks and towards better modeling of phenotypes. The methods we use can also be applied to networks from other disciplines.