Hubs with Network Motifs Organize Modularity Dynamically in the Protein-Protein Interaction Network of Yeast

Guangxu Jin,Shihua Zhang,Luonan Chen,Xiang-Sun Zhang,Nick Monk

doi:10.1371/journal.pone.0001207

Abstract

BackgroundIt has been recognized that modular organization pervades biological complexity. Based on network analysis, ‘party hubs’ and ‘date hubs’ were proposed to understand the basic principle of module organization of biomolecular networks. However, recent study on hubs has suggested that there is no clear evidence for coexistence of ‘party hubs’ and ‘date hubs’. Thus, an open question has been raised as to whether or not ‘party hubs’ and ‘date hubs’ truly exist in yeast interactome.MethodologyIn contrast to previous studies focusing on the partners of a hub or the individual proteins around the hub, our work aims to study the network motifs of a hub or interactions among individual proteins including the hub and its neighbors. Depending on the relationship between a hub's network motifs and protein complexes, we define two new types of hubs, ‘motif party hubs’ and ‘motif date hubs’, which have the same characteristics as the original ‘party hubs’ and ‘date hubs’ respectively. The network motifs of these two types of hubs display significantly different features in spatial distribution (or cellular localizations), co-expression in microarray data, controlling topological structure of network, and organizing modularity.ConclusionBy virtue of network motifs, we basically solved the open question about ‘party hubs’ and ‘date hubs’ which was raised by previous studies. Specifically, at the level of network motifs instead of individual proteins, we found two types of hubs, motif party hubs (mPHs) and motif date hubs (mDHs), whose network motifs display distinct characteristics on biological functions. In addition, in this paper we studied network motifs from a different viewpoint. That is, we show that a network motif should not be merely considered as an interaction pattern but be considered as an essential function unit in organizing modules of networks.

Highlights

Many types of molecular networks display scale-free topologies which are characterized by the power-law degree distribution [1,2,3,4,5]
We demonstrate that the network motifs of an motif party hubs (mPHs) are more likely to stay inside a protein complex with their mPH, control the local topological structure, locate in the same cellular localizations as the mPH, and co-express in microarray data
From the definition of mPHs and motif date hubs (mDHs), we can see that the network motifs of an mPH more likely stay together in the same protein complex as the mPH while those of an mDH spread outside the protein complex of the mDH

Summary

Introduction

Many types of molecular networks display scale-free topologies which are characterized by the power-law degree distribution [1,2,3,4,5]. ‘party hubs’ and ‘date hubs’, based on whether or not the hubs are co-expressed with their partners by using yeast microarray data [10]. Due to the topological difference between FYI and HCfyi, Batada et al found that there is no evidence for coexistence of party hubs and date hubs, and the results about ‘party hubs’ and ‘date hubs’ are totally not correct. The network motifs of these two types of hubs display significantly different features in spatial distribution (or cellular localizations), co-expression in microarray data, controlling topological structure of network, and organizing modularity. By virtue of network motifs, we basically solved the open question about ‘party hubs’ and ‘date hubs’ which was raised by previous studies. We show that a network motif should not be merely considered as an interaction pattern but be considered as an essential function unit in organizing modules of networks

Objectives

Methods

Results

Conclusion