Abstract
BackgroundUncovering the evolutionary principles of gene coexpression network is important for our understanding of the network topological property of new genes. However, most existing evolutionary models only considered the evolution of duplication genes and only based on the degree of genes, ignoring the other key topological properties. The evolutionary mechanism by which how are new genes integrated into the ancestral networks are not yet to be comprehensively characterized. Herein, based on the human ribonucleic acid-sequencing (RNA-seq) data, we develop a new evolutionary model of gene coexpression network which considers the evolutionary process of both duplication genes and de novo genes.ResultsBased on the human RNA-seq data, we construct a gene coexpression network consisting of 8061 genes and 638624 links. We find that there are 1394 duplication genes and 126 de novo genes in the network. Then based on human gene age data, we reproduce the evolutionary process of this gene coexpression network and develop a new evolutionary model. We find that the generation rates of duplication genes and de novo genes are approximately 3.58/Myr (Myr=Million year) and 0.31/Myr, respectively. Based on the average degree and coreness of parent genes, we find that the gene duplication is a random process. Eventually duplication genes only inherit 12.89% connections from their parent genes and the retained connections have a smaller edge betweenness. Moreover, we find that both duplication genes and de novo genes prefer to develop new interactions with genes which have a large degree and a large coreness. Our proposed model can generate an evolutionary network when the number of newly added genes or the length of evolutionary time is known.ConclusionsGene duplication and de novo genes are two dominant evolutionary forces in shaping the coexpression network. Both duplication genes and de novo genes develop new interactions through a “rich-gets-richer" mechanism in terms of degree and coreness. This mechanism leads to the scale-free property and hierarchical architecture of biomolecular network. The proposed model is able to construct a gene coexpression network with comprehensive biological characteristics.
Highlights
Uncovering the evolutionary principles of gene coexpression network is important for our understanding of the network topological property of new genes
According to the human gene age data, we reproduce the evolutionary process of this network and develop a new evolutionary model of gene coexpression network, which help us discover the evolutionary mechanism of biomolecular networks
This evolutionary mechanism satisfies the traditional “rich-gets-richer" mechanism for degree, and constrains the growth of genes with large degree but less important. This mechanism makes the evolutionary networks generating from our model are more similar to the real gene coexpression network. We find that they prefer to connect genes with large degree and large coreness
Summary
Uncovering the evolutionary principles of gene coexpression network is important for our understanding of the network topological property of new genes. Most existing evolutionary models only considered the evolution of duplication genes and only based on the degree of genes, ignoring the other key topological properties. Based on the human ribonucleic acid-sequencing (RNA-seq) data, we develop a new evolutionary model of gene coexpression network which considers the evolutionary process of both duplication genes and de novo genes. Recent studies have proposed some general properties of biomolecular networks. These properties can be summarized as follows: (1) The degree distribution of genes in the network obeys to the power law distribution [4] and the power exponent is between 1 and 2 [5,6,7]. The mean pathlength of the network is small and the average clustering coefficient is relatively large [10,11,12]. (4) These networks are sparse, which means the average number of edges connected to the gene is small [13]
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
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
