Fig. 3: Development of partner ensembles is CP-AMPAR-dependent.

A Representative identification of ensemble using PCA-ICA, with activity of highly weighted neurons (weights > mean + 2 SD) shown beneath. Ensembles consist of two or more neurons. B Total ensembles across days. Mixed LM treatment β = 0.023, p = 0.159, post-hoc Day 0 two-sample t(16) = –1.10, p = 0.293, Day 1 t(15) = –1.71, p = 0.127, Day 2 t(13) = –0.900, p = 0.388. C Partner-selective ensembles across days. Mixed LM treatment β = –0.022, p = 0.015, post-hoc day 0 two-sample t(16) = 0.491, p = 0.633, Day 1 t(15) = 1.153, p = 0.267, Day 2 t(13) = 2.535, p = 0.0262. D Partner-selective ensembles on Day 2. E Correlation in control animals between % time in partner chamber and partner minus stranger ensembles (divided by total # neurons). Spearman’s r(8) = 0.81, p = 0.0149. F Representative spatial clustering analysis of ensemble neurons. Highly weighted neurons in each ensemble were identified and average pairwise Euclidean distance was computed. G Ensemble clustering across animals. Quantified as a percentile of mean pairwise distances versus null distributions from random cell selections. One-sample t-test versus 50, vehicle t(7) = –2.63, p = 0.034, IEM-1460 t(2) = –11.139, p = 0.00796. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. Data are presented as mean values ± SEM.