Fig. 3: Individual PCs changed reference frames.
From: Mechanisms of experience-dependent place-cell referencing in hippocampal area CA1
a, An example of a space PC becoming goal-referenced. Heatmaps depict the spatial Δf/f activity across laps for day 1 on a familiar belt (left) and day 2 on a novel belt (right) during pre (top) and post (middle). White dashed lines indicate reward locations (pre, 50 cm; post, 140 cm). Peak-normalized spatial Δf/f during pre and post on day 1 (familiar, left) and day 2 (novel, right) (bottom). The heatmaps use smaller color scales (0–1) to visualize relatively low activity from individual cells. b, Same as a, but showing a PC that remained space-referenced on the same familiar belt for two consecutive days. c, Stability of space PCs across 2 days. The fraction of space PCs (day 1, familiar belt) that stayed space-referenced (dark blue) or switched to goal-referenced (dark red) on day 2 on a novel belt (left). The fraction of space PCs (day 1, familiar belt) that stayed space-referenced (light blue) or became goal-referenced (light red) on day 2 on the same familiar belt (right). d, Same as c, but showing the stability of goal PCs across 2 days. e, Resampled bootstrapped values of the ‘space/(space + goal)’ index across 2 days (F to F, familiar to familiar). The ‘space/(space + goal)’ index is calculated as the number of space PCs that stayed space-referenced divided by the total number of space PCs that stayed space-referenced or became goal-referenced on the next day. For each bootstrap, we randomly sampled a fraction of the real data with replacement to calculate the space/(space + goal) index using PCs on day 1. The dashed line represents the bottom fifth percentile of the bootstrapped values during F to F. The solid line represents the observed proportion from F to N (familiar to novel). f, Same as e, but showing the stability of goal PCs across 2 days. n = 6 mice for all conditions shown in c and d.
