Fig. 3: Backbone sequences have rigid and flexible elements.
From: Backbone spiking sequence as a basis for preplay, replay, and default states in human cortex
a Representative backbone sequence from a given participant (left) and two separate units are shown with individual traces indicating the cross-correlations of the spike train of that unit with the summed spike trains of all other units in the population (right). The rigid unit has a highly consistent temporal spiking relationship with the population activity whereas the flexible unit is much more inconsistent. Darker and lighter lines indicate earlier and later in the behavioral session, respectively. b Corresponding histograms of the normalized sequence rank for each 2 min bin for each unit across the entire recording session. c Relationship between Z-scored normalized sequence rank variance and firing rate for all units from a representative session. More negative values along the x-axis indicate a higher degree of rigidity. Each blue dot represents one isolated unit from this session. d Rigid units demonstrated significantly higher firing rates than flexible units across all participants (N = 6 participants, two-sided t-test, t(5) = 5.58, p = 0.003). (**) indicates p < 0.01. Error bars indicate SEM across participants.
