The fraction of interparalog edges for the extended DD model under different parameter conditions. As is the case for ΔD values, an increase in values of p and q and decrease of a lead to higher fraction of edges that connect paralogs in the network, divergence symmetry s has very little effect. Note that low values of p already have a large impact on the fraction of edges connecting paralogs: with p = 0.2 and a = 0.2, the probability of gaining an edge between daughter nodes after a duplication event in the network is p*(1-a) = 0.16, yet this small probability of an edge gain between daughter nodes after duplication leads to ~30% of all edges connecting paralogs. Default parameter conditions are p = 0.2, q = 0.7, s = 0.5, a = 0.2, each plot shows the fraction of interparalog edges when one of these parameters is varied while the others are kept at default values (for full parameter sweeps we refer to Figure S18). The gray line is the fraction of edges that connect paralogs of the yeast LC PIN. Boxes show the .25 and .75 percentile of 20 runs, the error bars show the extreme values and the black line is the mean of 20 runs.