EGN: a wizard for construction of gene and genome similarity networks

Abstract

BackgroundIncreasingly, similarity networks are being used for evolutionary analyses of molecular datasets. These networks are very useful, in particular for the analysis of gene sharing, lateral gene transfer and for the detection of distant homologs. Currently, such analyses require some computer programming skills due to the limited availability of user-friendly freely distributed software. Consequently, although appealing, the construction and analyses of these networks remain less familiar to biologists than do phylogenetic approaches.ResultsIn order to ease the use of similarity networks in the community of evolutionary biologists, we introduce a software program, EGN, that runs under Linux or MacOSX. EGN automates the reconstruction of gene and genome networks from nucleic and proteic sequences. EGN also implements statistics describing genetic diversity in these samples, for various user-defined thresholds of similarities. In the interest of studying the complexity of evolutionary processes affecting microbial evolution, we applied EGN to a dataset of 571,044 proteic sequences from the three domains of life and from mobile elements. We observed that, in Borrelia, plasmids play a different role than in most other eubacteria. Rather than being genetic couriers involved in lateral gene transfer, Borrelia’s plasmids and their genes act as private genetic goods, that contribute to the creation of genetic diversity within their parasitic hosts.ConclusionEGN can be used for constructing, analyzing, and mining molecular datasets in evolutionary studies. The program can help increase our knowledge of the processes through which genes from distinct sources and/or from multiple genomes co-evolve in lineages of cellular organisms.

Metatranscriptome analyses indicate resource partitioning between diatoms in the field

“…These translated peptide sequences were then combined with the translated proteins from the diatom genomes Fragilariopsis cylindrus CCMP1102 version 1.0, P. tricornutum CCMP632 version 2.0, Pseudonitzschia multiseries CLN-47 version 1.0, and T. pseudonana CCMP1335 version 3.0, which were collected from the Joint Genome Institute database (genome.jgi-psf.org/). A protein similarity network was then created using EGN, a software program that automates the reconstruction of gene networks from protein sequences through reciprocal BLASTp analysis (e-value <1e-5, 20% hit identity threshold, 5% best reciprocal threshold of best e-value, 90% minimal match coverage threshold) (80,81). Networks were then visualized and manipulated using Cytoscape 3.0, where the layout of the network was produced using an edge-weighted, spring-embedded model based on e-value, meaning that genes that are closer together are more similar (82,83).…”

Mitochondrial Genome Rearrangements in Glomus Species Triggered by Homologous Recombination between Distinct mtDNA Haplotypes

“…Two similarity network analyses, where each node represents a dpo sequence and an edge represents a sequence similarity/identity were performed using EGN (Halary et al 2013). The first network was constructed from a data set of 91 Glomeromycotan dpo sequences (from our collection and retrieved from GenBank).…”

Origin and Evolution of the Self-Organizing Cytoskeleton in the Network of Eukaryotic Organelles

“…Cluster maps can also provide a general overview of taxonomic distribution and sequence similarity, parameters that are not easily inferred from phylogenetic trees. Clustering is best thought of as a representation of sequence data as a similarity network, allowing evolutionary biologists to draw inferences about sequence evolution that are complementary to answers based on phylogeny (for a thoughtful introduction to the use of similarity networks, see Halary et al 2013).…”