Figure 1. Transient changes in the basal ganglia output are reduced by ongoing population activity.
A data set of 599 single units was isolated from recording sessions with at least eight simultaneously recorded units (Pan et al., 2013). Electrodes were targeted to the substantia nigra and ventral tegmental region of mice trained to perform an auditory trace conditioning paradigm. (A) Spiking activity was z-scored, aligned to the onset of the conditioned stimulus (CS), and averaged for all units. (B) For each recorded unit the mean subtracted response (RESPsingle) was computed as a function of the mean normalized activity prior to CS onset for the rest of the simultaneously recorded population (PREpopulation; 7–20 units). Population data were binned, averaged, and fit with a sigmoid function (cyan line). (C) The correlation coefficient between RESPsingle and PREpopulation was computed for each session (n = 32). A histogram of all correlation scores is drawn with significant correlations (permutation test) indicated by filled gray bars. The correlation score of the entire population (−0.1) is indicated by a cyan triangle.