Bayesian differential analysis of gene regulatory networks exploiting genetic perturbations

摘要

GRNs visually reflect the gene-gene interactions, which are significant for understanding gene functions and biological activities. In the past few years, a series of inference algorithms have been proposed to reconstruct topology structures of GRNs. Some computational methods were only developed to infer GRNs from gene expression data, such as Boolean networks [ ], mutual information models [ , ], Gaussian Graphical models [ , ], Bayesian networks [ , ] and linear regression models [ , ]; several other methods were also built to improve the accuracy of inference by integrating genetic perturbations with gene expression data, among which the algorithms based on SEMs [ – ] are one of the most popular approaches.