Fig. 2
From: Concurrence of form and function in developing networks and its role in synaptic pruning
Synaptic pruning. Experimental data sets on connectivity during infancy (points) and model fit (solid lines). a Data points correspond to synaptic density, ρ(t)∝κ(t), in the human infant brain, obtained in autopsies by directly counting synapses in tissues from different layers of the auditory cortex (here shown layers 1 (L1) and 2 (L2))32. The solid lines present the best fit obtained by the model, with τp,L1 = 1600(300) and τp,L2 = 3800(100). Other parameters are extracted from the data: L1: κ0 = 59.1, κ∞ = 30.7 and t0 = 700; and L2: κ0 = 75.1, κ∞ = 40.8 and t0 = 700, where t0 is set at the onset of pruning (i.e., corresponding to the maximum of κ(t)). Inset on the right shows the fit of the maximum on a log–log scale, labels as in the main plot. Parameters from the fit are aL1 = 33(5), τg,L1 = 210(20), τp,L1 = 3800(200), aL2 = 1.2(2), τg,L2 = 2700(100), τp,L2 = 290(40). The three lower points for t ≈ 5000 have been excluded from the fits. b Data points correspond to synaptic density measured via large-scale brain imaging experiments that quantify the number of connections in the developing mouse somatosensory cortex21. In blue solid lines (P1), fit with the linear model, for parameters τ p = 5.72(1), κ0 = 3.10, κ∞ = 1.64 and t0 = 18.97. In green solid lines (P2), fit including the growth factor, with parameters a = 1.7(1), τ g = 4.0(2) and τ p = 2.2(2). The third point form the right has been excluded from the fits. Error bars of the data points correspond to s.d. as obtained in the original works
