Fig. 3: Sustaining a cellular network.
From: Sustaining a network by controlling a fraction of nodes

a We apply our framework to the regulatory dynamics captured by Michaelis–Menten model44. b Simulations (symbols) and theory (lines, Eq. (10) with ρ = 0) results for a free system of Erdős-Rényi (ER) structure with N = 104 and κ = 40 and for the parameters’ values a = 1 and h = 2. There is a bi-stable region (gray shade). In case that a ≥ h, there is no bi-stability38, thus here, we consider h > a to analyze system sustaining. c Demonstration of the activities of the system with β = 3.1 for ρ = 0. Both states are stable, therefore the system is unsustainable. d For a forced system by a fraction ρ = 0.03 of random nodes with activity Δ = 5, Eq. (8) provides a phase diagram (thick curve), exhibiting an s-shape curve which has now also a regime with only a single active state (blue shade). This regime is a sustainable phase. Note that simulations (symbols) are in agreement with the theory (lines). The network is the same as in (b). e The same as (d) with a larger fraction of controlled nodes, ρ = 0.11. Here we see that the unsustainable region almost vanishes, and the transition becomes almost continuous. This agrees with Eq. (14). f Activities for β = 1.5, 2.5, 3.1 in three regions, inactive (red diamond), unsustainable (cyan circle), and sustainable (blue triangle) correspondingly. The dark blue nodes are the forced nodes. The red nodes represent x0, the cyan nodes represent unsustainable x1, and the light blue nodes represent sustainable x1. g The new phase diagram in (β, ρ)-space for Δ = 5. The simulations were done on ER networks with N = 104 for 50 values of ρ, 50 values of β, for κ = 20, 60, 100, and averaged over ten realizations. In the color bar, the value 0 represents an unsustainable system, and 1 represents the other cases. The black lines are obtained from Eqs. (10) and (11). h The same as (g) with lower intervention force Δ = 1. As expected, in this case, a larger fraction of controlled nodes is needed to make the network sustainable. i (κ, λ)-space for a free system. j The sustaining phase diagram for ρ = 0.01 and Δ = 5. The light blue is the sustainable phase when forcing a fraction ρ = 0.01 of nodes, and the dark blue is the sustainable regime for holding a single node. The white lines represent the theory, Eqs. (10) and (11). k Horizontal trajectory in (κ, λ)-space for fixed λ = 0.1 and varying κ. Symbols are simulations and the line is theory obtained from Eqs. (10) and (11). The slope is according to Eq. (12). l Vertical trajectory in (κ, λ)-space for fixed κ = 20, 100 and varying λ. Note that the critical fraction for sustaining, ρc, for a given λ approaches 0, where it reaches the single-node sustainable phase in simulations (symbols). The theory, Eqs. (10) and (11) (continuous lines), deviate from the simulation results for small ρ. The dashed lines are from a different theory, see Supplementary Note 3, which captures also the limit of small ρ.