Abstract
ABSTRACTThe gene neighborhood in prokaryotic genomes has been effectively utilized in inferring co-functional networks in various organisms. Previously, such genomic context information has been sought among completely assembled prokaryotic genomes. Here, we present a method to infer functional gene networks according to the gene neighborhood in metagenome contigs, which are incompletely assembled genomic fragments. Given that the amount of metagenome sequence data has now surpassed that of completely assembled prokaryotic genomes in the public domain, we expect benefits of inferring networks by the metagenome-based gene neighborhood. We generated co-functional networks for diverse taxonomical species using metagenomics contigs derived from the human microbiome and the ocean microbiome. We found that the networks based on the metagenome gene neighborhood outperformed those based on 1748 completely assembled prokaryotic genomes. We also demonstrated that the metagenome-based gene neighborhood could predict genes related to virulence-associated phenotypes in a bacterial pathogen, indicating that metagenome-based functional links could be sufficiently predictive for some phenotypes of medical importance. Owing to the exponential growth of metagenome sequence data in public repositories, metagenome-based inference of co-functional networks will facilitate understanding of gene functions and pathways in diverse species.
Highlights
The functional and regulatory constraints acting on genes have contributed to shaping pathways and functional modules
We demonstrated that the metagenome-based gene neighborhood could predict genes related to virulence-associated phenotypes in a bacterial pathogen, indicating that metagenome-based functional links could be sufficiently predictive for some phenotypes of medical importance
The overall procedure for the construction of the metagenome-based gene neighborhood network is summarized in Figure 1 as briefly described in the Materials and Method section
Summary
The functional and regulatory constraints acting on genes have contributed to shaping pathways and functional modules. Large-scale identification of genes contributing to the operon structure generally relies on measurement of the chromosomal distance between genes (Dandekar et al 1998, Overbeek et al 1999) or determining the probability that a given pair of conserved genes is in close proximity to each other across prokaryotic genomes (Bowers et al 2004). These two approaches for the inference of functional links based on gene neighborhoods are complementary (Shin et al 2014). Identifying the gene neighborhood can help to infer co-functional links in prokaryotic species and in higher eukaryotes using orthologous gene neighborhoods in prokaryotic genomes
Sign-in/Register to view highlights & summary 
Published Version (
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