The reconstruction of gene regulatory network based On Multi-Agent System by fusing multiple data sources

Gene regulatory network (GRN) is a very important biological system during the cell cycle. In this case, the gene regulatory network reconstruction is an important and meaningful work. Based on the network, the future state of a cell can be predicted by the gene expression process. In this paper, we present a new method to reconstruct the network. During this method, we use Multi-Agent System (MAS) to fuse the gene expression data and TF binding data and generate an initial network. Based on the initial network, a final network is learned using Dynamic Bayesian Network (DBN) learning method. In order to verify the performance of our method, we experiment the method using the data of 25 genes and compare the result with the algorithms already raised in the previous papers. The comparing result show that the method based on MAS and DBN has a better performance than others.