Detection of composite communities in multiplex biological networks.

Laura Bennett,Sophia Tsoka,Aristotelis Kittas,Gareth Muirhead,Lazaros G Papageorgiou

doi:10.1038/srep10345

Abstract

The detection of community structure is a widely accepted means of investigating the principles governing biological systems. Recent efforts are exploring ways in which multiple data sources can be integrated to generate a more comprehensive model of cellular interactions, leading to the detection of more biologically relevant communities. In this work, we propose a mathematical programming model to cluster multiplex biological networks, i.e. multiple network slices, each with a different interaction type, to determine a single representative partition of composite communities. Our method, known as SimMod, is evaluated through its application to yeast networks of physical, genetic and co-expression interactions. A comparative analysis involving partitions of the individual networks, partitions of aggregated networks and partitions generated by similar methods from the literature highlights the ability of SimMod to identify functionally enriched modules. It is further shown that SimMod offers enhanced results when compared to existing approaches without the need to train on known cellular interactions.

Highlights

Cellular organisation is assumed to be modular[1], with each module driving a distinct biological process
Community structure detection has been explored within the context of multiplex networks, i.e. networks with edges that are categorised according to type, sometimes known as multi-dimensional, multi-layer or multi-slice networks[9], where each edge type is associated to an individual network slice or layer
We report a mixed integer non-linear programming (MINLP) model, SimMod, which takes multiple network slices as input, optimises average modularity across all slices and returns a single partition of composite communities

Summary

Introduction

Cellular organisation is assumed to be modular[1], with each module driving a distinct biological process. This topology is known as community structure[2] and its detection is widely accepted as a means of revealing the relationship between topological and functional features of biological systems[3]. Communities, known as modules, have been shown to comprise groups of biomolecules that physically interact, are functionally cohesive, co-regulated or correspond to biological pathways[4]. Applications of community structure detection to biological systems often consider networks of a single interaction type. With regards to existing methods that target composite module detection, two models have been proposed to derive composite modules from physical and genetic interactions. Similar methods are described elsewhere[13,14]

Objectives

Methods

Results

Conclusion