Generating influence networks. Influence networks are directed weighted networks (right) generated from undirected weighted networks (left). Illustrated here is an example of a coexpression network and its influence network. Influence is calculated as the proportion of total edge weights a given node shares with each of its neighboring nodes (see Methods). Here, red links show the influence of node A on nodes B and C, which are completely influenced by node A because they do not have other neighbors to influence them. Blue links show the influence of nodes B and C on node A. Node B has a greater influence on node A than node C because the weight of the edge linking B to A is greater than the weight of the edge linking C to A. The cumulative influence of each node is calculated by summing the values of all outgoing edge weights from a given node.