Response inhibition in premotor cortex corresponds to a complex reshuffle of the mesoscopic information network

Abstract Recent studies have explored functional and effective neural networks in animal models; however, the dynamics of information propagation among functional modules under cognitive control remain largely unknown. Here, we addressed the issue using transfer entropy and graph theory methods on mesoscopic neural activities recorded in the dorsal premotor cortex of rhesus monkeys. We focused our study on the decision time of a Stop-signal task, looking for patterns in the network configuration that could influence motor plan maturation when the Stop signal is provided. When comparing trials with successful inhibition to those with generated movement, the nodes of the network resulted organized into four clusters, hierarchically arranged, and distinctly involved in information transfer. Interestingly, the hierarchies and the strength of information transmission between clusters varied throughout the task, distinguishing between generated movements and canceled ones and corresponding to measurable levels of network complexity. Our results suggest a putative mechanism for motor inhibition in premotor cortex: a topological reshuffle of the information exchanged among ensembles of neurons.

and V D out (OUT markers) distribution parameters (x axis: mean, µ; y axis: variance, σ 2 ) of the PMd empirical information network across behavioral conditions for each recording session for both monkeys.As evident, only the V D out distributions showed a great excursion of the variance with respect to the mean, i.e. show the presence of a fat-tail.Indeed, for the V D out distributions σ 2 is at least one order of magnitude greater than µ (Kruskal-Wallis, p < 0.01).This documents the presence of a fat tail and hence of high out degree nodes.On the contrary, the V D in distributions showed comparable values.
Figure S1.V D in(IN markers)  and V D out (OUT markers) distribution parameters (x axis: mean, µ; y axis: variance, σ 2 ) of the PMd empirical information network across behavioral conditions for each recording session for both monkeys.As evident, only the V D out distributions showed a great excursion of the variance with respect to the mean, i.e. show the presence of a fat-tail.Indeed, for the V D out distributions σ 2 is at least one order of magnitude greater than µ (Kruskal-Wallis, p < 0.01).This documents the presence of a fat tail and hence of high out degree nodes.On the contrary, the V D in distributions showed comparable values.

Figure S2 .Figure S3 .Figure S4 .Figure S5 .
Figure S2.Comparison between the V D out distribution parameters (x axis: mean, µ; y axis: variance, σ 2 ) of the PMd empirical information network across behavioral conditions for each recording session for both monkeys and the ensemble-average obtained from the null model (see text for details).The empirical distributions have a variance that spans values at least one order of magnitude greater than that of the networks derived from the MUA distribution, despite having comparable µ values; Kruskal-Wallis, p < 0.01, Bonferroni corrected for multiple comparisons.(NM: Null model).

Table S1 .
Behavioural results.S, index of the recording session.RT Go , mean reaction time of Go trials.RT W r , mean reaction time of wrong Stop trials.SSD CS , mean SSD of correct Stop trials.SSD W r , mean SSD of Wrong Stop trials.SSRT , Stop signal reaction time.P inhibit , inhibition probability.The p-values result from the independence; Kolmogorov-Smirnov test, p < 0.05 between RT Go and RT W r distributions.(Go: Go trials; WS: wrong Stop trials; CS: correct Stop trials)

Table S2 .
I matrix details.Values of I for each cluster and each behavioral condition averaged (mean ± SEM) over recording sessions.The symbol / marks the absence of self loops which are excluded from the analysis (see main text). .

Table S4 .
ANOVA results.S, index of the recording session.N Go−CS , number of modules with p < 0.01 between Go and correct Stop trials.N Go−W S , number of modules with p < 0.01 between Go and wrong Stop trials.N CS−W S , number of modules with with p < 0.01 between correct and wrong Stop trials.All Ns are in percentage over the total number of moduels.p-values are Bonferroni corrected for multiple comparisons.(Go: Go trials; WS: wrong Stop trials; CS: correct Stop trials)

Table S5 .
Clusters composition.For each monkey the composition of clusters averaged (mean ± SEM) over recording sessions is reported.N is the number of modules available.TableS6Details of V D out across behavioral conditions.For each monkey V D out values averaged (mean ± SEM) over recording sessions are reported.The cluster_3 showed the highest values of V D out compared to other clusters in all behavioural conditions.The cluster_1 and the cluster_2 showed the second highest V D out values during both Go and wrong Stop trials and correct Stop trials respectively; Kruskal-Wallis, p < 0.01 Bonferroni corrected for multiple comparisons.N: number of session available.