Fig. 6: Cell assembly properties and spatial coding abilities in early development.

a An example of raster plots showing cell assemblies detected from all cells recorded in a session on the linear track at P21. These cell assemblies were preordered by a K-means clustering algorithm (see Methods). b Correlation matrix of the activities of all cells from the same session as in (a), with cells organized according to the cell assemblies identified in (a). c Detection of synchronous calcium events (SCEs) in an assembly. Top, the raster plot of an example assembly recorded at P21; bottom, the histogram of the sum of the calcium events as in top. The shaded red vertical areas at the top and the red bars at the bottom mark the detected SCEs. d Correlation matrix for all cells during a linear track session at P18, with cells organized according to the cell assemblies. e Top: raster plots of three example cell assemblies from (d). Neurons within the same assembly were labeled in the same color. Bottom: count density curves illustrating the normalized counts of calcium events in each bin within an assembly. f, g Same format as (d) and (e), but for a session recorded at P25. h, i The synchronization indices for all cells at different developmental ages (h) and adults (i). Dots and error bars indicate mean ± SEM. The indices increased with age (Pearson correlation coefficient r = 0.26, p < 10 ^− 15, n = 47138 cells), but not with recording days in adults (r = −0.1, p = 3.48 × 10^–13, n = 10526 cells). j, k The relationship between the synchronization index of the cells and their wSI ranks at various developmental ages (j) and adults (k). Cells were divided into ten equal groups according to their wSI scores at each age. A linear regression fit was applied across these ten groups. Dots with bars indicate mean ± SEM. The r² values and p values for the fit were labeled at the bottom. Source data are provided as a Source Data file.