Fig. 5: Psilocybin leads to decreased mean square displacement (MSD) of neural spiking activity and decreased Lempel-Ziv complexity (LZC).

A The spiking activities of recorded single units in the mPFC can be used to describe the binary state transitions of a representative network of the units: (i, left) An example three-neuron network is shown, with possible states visualised as the vertices of a unit cube. The possible 3-bit binary states, where 0 (light-orange) represents no spikes in the time bin, and 1 (dark-orange) represents spiking. The axes represent the binary values of the individual neurons. The neural network (dark-grey ball) transitions from one state to another with time. (i, right) The MSD is a Euclidean jump distance between any two network states over a discrete time-step ∆t. (ii) The temporal dynamics of the network are shown as consecutive 3-bit states at discrete time points, t₁ to t_n. At every time point, an average MSD of the network transitions over the next 1 s is computed, giving a distribution of MSDs characterising network transitions across the entire time duration. MSD distributions centred on lower values reflect restricted state transitions, while higher values indicate relatively free network state transitions. (iii) Schematic descriptions of the potential energy landscapes, E_p, for the low and high MSD distributions are shown. Shallower landscapes (left) require less activation energy leading to more frequent state transitions (light grey arrow) and a high MSD distribution. Network states sitting in a deep basin of attraction (right) requires larger activation energy, E_a, to jump to neighbouring states, causing less-frequent transitions (dark grey arrow) and low MSD distribution. B MSD from all single units across pre- and post-injection rest blocks for each drug condition, indicating a decrease in MSD under psilocybin. C Change in MSD from pre- to post-injection rest blocks (left) and task blocks (right), across drug conditions, indicating a dose-dependent decrease in MSD with psilocybin. D LZC from all single units across pre- and post-injection rest blocks for each drug condition. E Change in LZC from pre- to post-injection rest blocks (left) and task blocks (right), across drug conditions, indicating a decrease in LZ complexity with psilocybin (0.3 and 1 mg/kg) in both rest and task contexts. Error bars denote SEM (N = 4). Asterisks denote significant post-hoc t-test Bonferroni corrected differences between saline and 0.3 mg/kg or 1 mg/kg psilocybin; ***p < 0.001, •p < 0.05 but did not survive Bonferroni multiple comparison correction.