Extended Data Fig. 2: Effect of CFA inactivation on muscle activity.

a, Muscle activity times series surrounding three inactivation trials in one mouse. The activity of each muscle was z-scored, so units are standard deviations of the original recorded signal. Vertical cyan bars in a, b, e–g indicate the 25 ms epoch of blue light applied to CFA. Gray dotted lines are 10 ms after light onset, the shortest latency at which effects can be observed. b, Mean ± sem muscle activity for control (gray) and inactivation (cyan) trials for the recorded forelimb muscles in three additional mice. c, Absolute difference between inactivation and control trials at 25, 50, and 100 ms after trial onset for 8 individual mice (black circles) and the mean across mice (red bars). The mean was significantly greater than 0 at 25, 50 and 100 ms following light onset (p = 0.004 in each case, two-sided Wilcoxon signed-rank test). Despite this, we found that the covariance of muscle activity patterns was not substantially different in the 50 ms following control and inactivation trial onsets d, Mean ± sem (n = 8 mice) variance capture of muscle activity during inactivation trials and one half of control trials using principal components computed from the other half of control trials. Only time series during the 50 ms following trial onset were used here to focus on the period when inactivation effects were most apparent. The covariance of muscle activity patterns was not substantially different in the 50 ms following control and inactivation trial onsets. e–g, Mean ± sem muscle activity for control (gray) and inactivation (cyan) trials using the first 10 and last 10 trials of each type from each session (e), the first 13 and next 13 sessions for each mouse (f, n = 5 mice), and the first 500/1000 inactivation/control trials and the next 500/1000 inactivation/control trials for each mouse (g, n = 6 mice). Average inactivation effects on muscle activity show remarkable consistency, both within and across sessions.