Generating time series simulation dataset derived from dynamic time-varying Bayesian network

Abstract

Numerous network inference models have been developed for understanding genetic regulatory mechanisms such as gene transcription and protein synthesis. Dynamic Bayesian network effectively represent the causal relationship between genes and gene and protein. Modern approaches employ single multivariate gene expression data set to estimate time varying dynamic Bayesian network. However, evaluating inferred time varying network is infeasible due to the absence of known gold standards. In this paper, the simulation model for time series gene expression level under certain network structure is proposed. The network can be used for assessing inferred data which is estimated based on simulated gene expression data.