Extended Data Fig. 3: pPCx location selectivity remains stable across recording blocks.

(a) Rat position on the maze during two blocks within the same recording session. (b-d). Occupancy normalized firing rate heat map, PETH, and raster plots associated with example cells 1-3 in Fig 2a–c. PETH and raster plots were aligned to initiation port poke-in and sorted by location. (e). Z-scored mean firing rate of neurons for different port locations across blocks. (f) Stability of port location representation between the two blocks was calculated the pairwise Pearson’s correlation coefficient for population activity across different locations in these two blocks. Values along the diagonal band in the correlation coefficient matrix represent the similarity of the pPCx population response to the same location in two recording blocks. (f, top) Similarity matrix obtained by computing the pairwise Pearson’s correlation coefficients between locations population response vectors for blocks 1 and 2 of example session shown in (a-e). Coefficients along the diagonal band are significantly higher than off-diagonal (r_diag = 0.80 ± 0.05, r_offdiag = 0.54 ± 0.09, Wilcoxon rank-sum test, p < 0.01), indicating that location representations were stable between blocks. (f, bottom). Same analysis as in the top panel but for all pPCx neurons recorded (r_diag = 0.81 ± 0.02, r_offdiag = 0.47 ± 0.02, Wilcoxon rank sum test, p < 0.001, n = 44 sessions, 995 neurons). We conclude from these data that uncontrolled odours were unlikely to account for the pPCx location representations observed.