Fig. 3
From: Constructing cancer patient-specific and group-specific gene networks with multi-omics data
Process of constructing a group-specific gene network from multiple patient-specific gene networks. In the group-specific gene network, the node size and node color are proportional to the average degree and average \(\left | \Delta PCC \right |\) of the node, respectively. If the dominant type for a particular edge is ‘correlation-gained’ (positive ΔPCC) in the patient-specific networks, the edge is represented in red in the group-specific network. If the dominant type for a particular edge is ‘correlation-lost’ (negative ΔPCC) in the patient-specific networks, the edge is represented in blue in the group-specific network