Extended Data Figure 9. Simulation of input convergence in Dbh-Cre tracing experiments.

In the sparsest Dbh-Cre trans-synaptic tracing brain, 4 starter cells received input from 43 distinct input regions (309 input neurons, see Supplementary Table 2, Sample #8). In the second sparsest sample, 22 starter cells received input from 66 distinct input regions (756 input neurons; see Supplementary Table 2, Sample #9). a, The relation between the number of input regions for each LC-NE starter cell and the probability of observing > 42 (left) or > 65 (right) input regions in simulation, assuming that each starter cell receives input from a given region with the same probability. As the number of input regions per starter cell increases, the probability of observing inputs from > 42 or > 65 regions also increases. Based on a threshold of p value < 0.001, these simulations suggest that, to account for the total number of observed input areas in each brain sample, there must be individual LC-NE neurons that receive input from more than 15 regions for the sparsest sample (red dot, left) or more than 9 regions for 2^nd sparsest sample (red dot, right). b, Detailed view of the distribution of simulation results corresponding to the red dots in (a). Assuming that each cell receives input from 15 (left) or 9 (right) distinct regions, only 5 (left) or 6 (right) out of 10,000 simulations label > 42 (left) or > 65 (right) input regions. Note that if the assumption that each starter cell receives input from the same number of regions does not apply, then there must be at least one cell receiving input from more regions than the number specified in the simulation.