Optimized time-lag differential method for constructing gene regulatory network

Abstract

One major molecular genetics control mechanism is transcriptional gene regulation where regulatory genes control other genes over time. The biological approaches to determine the Gene Regulatory Networks (GRN), require costly experimental setup and time for generating voluminous time series gene expression data. Computational model has been developed to generate GRN by analyzing time series gene expression data. Sometimes gene to gene interaction not performing in order due to change of activity in the cell, resulting regulatory error, creating difficulty to identify the target genes. Searching of regulatory genes with permissible error, producing maximum number of target genes is the real challenge for predicting the most important and instant regulating pathway of the human GRNs. The aim of the paper is to establish a computational methodology to reconstruct the GRN from short time series gene expression data with an objective to explain the observed behavior of biological systems. We employ nondominated sorting genetic algorithm (NSGA-II) to develop the proposed Optimized time-lag differential (OTD) algorithm, which identifies maximum number of target genes for the selected regulatory genes within minimum regulating time. Experiments have been conducted using eight different time series microarray data set showing better performance compare to others.