Extended Data Fig. 8: Intersection with grid cells, head-direction cells and boundary-vector cells.
a, Colour-coded rate maps showing 12 representative grid cells recorded in the absence (left) and presence (right) of objects on consecutive trials. Mouse and cell ID numbers are indicated to the left of each pair of columns; grid score (G.s.) and peak rate are shown below each rate map. Scale bar indicates firing rate. Cells in the top three rows of the left pair of columns are examples of grid cells that expressed an extra field in the vicinity of at least one of the objects. Such effects were observed in 10 out of 124 grid cells. Only four grid cells passed the criteria for object-vector cells. Grid fields from the no-object trial mostly retained their firing locations when the object was added, but in a few cases single fields were moderately displaced or the grid underwent mild disruption. b, Colour-coded rate maps and head-direction tuning curves for six representative head-direction cells recorded with no object (left) and with an object (right) on consecutive trials. Mouse and cell ID numbers are indicated to the left; peak rate is shown below each rate map and head-direction score and peak rate are shown below each head-direction tuning curve. Sharply tuned head-direction cells mostly failed to develop vector fields in the presence of objects (exemplified by the cells in the top three rows). One of the few object-vector cells that also had sharp head-direction tuning is cell 398 in the fourth row. The majority of object-vector cells that passed as head-direction cells had only moderate head-direction tuning, and this tuning was clearly reduced when the object was introduced (Extended Data Fig. 4c, d and exemplified by cells 351 and 266 in the two bottom rows of this panel). c, Colour-coded rate maps for all cells that pass a criterion for boundary-vector cells. Symbols as in a. Twenty out of 840 cells passed the criterion for boundary-vector cells. Among these, only one had a firing field that did not encroach upon the wall (mouse 80569, cell 524, highlighted in square frame). d, Distance between boundary and centre of boundary-vector firing field (n = 20 fields from 20 cells) for all cells that passed the criterion for boundary-vector cells, compared to distance between nearest point of object and centre of object-vector fields (n = 221 object-vector fields from 162 cells). Black line between box edges indicates median, box edges indicate 25th and 75th percentiles, whiskers extend to the most extreme point that lies within 1.5 × IQR, and data points larger than 1.5 × IQR are considered outliers (red crosses). The distribution of distances in object-vector fields is skewed towards greater distances than the distribution of distances in boundary-vector fields (two-sided Mann–Whitney U-test, U = 703, P = 9.0 × 10−9).
