Network derived from the domain architectures of toxins. The central panel shows the network for all toxins in all species, whereas the lower panels show networks derived for major bacterial clades. The network is a directional graph with edges connecting neighboring domains in a polypeptide, in which the N-terminal domain is the source node, whereas the C-terminal domain is the target node. Edges are colored to match the source node color to illustrate the main direction of flow in the graph. Domains with similar properties are grouped together as shown.