Fig. 1

A small network with two genes. Gene s1 has a negative feedback loop on itself, gene s2 has a negative feedback loop on itself and a positive feedback on gene s1. a The network model. There are three binding sites, represented by triangles (b1, b2, b3), two control functions, represented by rectangles, and two substance generators, represented by diamonds (s1, s2). Discrete inputs and outputs correspond to dotted lines, while continuous inputs and outputs are represented by continuous lines. For each binding site, the association constant is given above and the dissociation constant below an incoming arrow. The substance generators s1 and s2 can either produce a substance or not – here linear growth and degradation rates are written above and below the output line. The binding sites can be unoccupied (‘0’ in the table) or occupied (‘1’ in the table). b A simulation run of the network. With association and dissociation constants given in (a) the network switches between two modes where either s1 is active and s2 inactive (↑↓) or vice versa (↓↑). c A graphical representation of the qualitative observational sequence describing the simulation run (b) of the network given in (a).