Fig. 2
The flowchart of model architecture and the structure of the paper. The model takes microarray data as input, and will be trained as described in flowchart: finding out the relationship between any one gene and other n − 1 genes, making adjacency matrix, building gene regulatory network and getting the final gene network according to the weight ratio λ. The training is carried on in each group respectively. The network is compared with the common relevant network by the value of parameters and the differential genes determined by the network are compared with that determined by fold_change