Multifactorial Modelling of MicroRNA Associated Repression and Its Subsequent Effects on Gene Expression in MicroRNA:Target Network
Abstract
MicroRNAs negatively regulate expression of many genes, efficiency of which depends on concentration of targets, content and structure of seed region. Current models consider one microRNA and its targets, or one mRNA and microRNAs targeting that mRNA. In this study, a network-based model was developed incorporating factors that are important in microRNA activity such as free energy, microRNA expression and gene expression levels, seed structure and position of target region on mRNA. The gene and microRNA expression data were downloaded from The Cancer Genome Atlas (TCGA), microRNA:target pairing data was obtained from previously performed high-throughput sequencing studies using CLASH and CLEAR-CLIP. In this regard, microRNA expression, gene expression and microRNA:target databases were combined and the initial network created from the dataset was accepted as steady-state. The model was used to calculate how the expression of other genes will change in the network upon perturbation of single gene expression. As an example, in a microRNA:target network extracted from a breast cancer patient with 61 microRNAs and 186 genes, two-fold increase in one of the genes resulted in 15% of targets being up-regulated and 81% being constant . When gene expression changes calculated for the genes two genes node away from initial perturbed gene, we observed increase in 53% and decrease in 18% of all genes in the network. Our model can help understand gene expression changes in context of complex microRNA:target network and pave the way for gene expression analysis in context of ceRNAs such as circRNAs, lncRNAs.
