Fig. 3: IL^→NAcSh neurons encode a generalized social identity of familiar conspecifics.

a Subject mice interacted with two littermates (L1 and L2) for 5 min each in two separate sessions with a 5-min interval. b Subject mice showed significantly reduced interaction times. One-way RM ANOVA: n = 13 mice; F_3,36 = 5.113, **p = 0.0048. c Subject mice interacted with two novel conspecifics (N1 and N2) for 5 min in two sessions with a 5-min interval. d Subject mice interacting with novel conspecifics showed significantly reduced interaction times. One-way RM ANOVA: n = 13 mice; F_{2.006, 24.07} = 13.29, ***p = 0.0001. e Identity decoding using a CCGP analysis of the consecutive littermate recognition task data. f A representative image showing that IL^→NAcSh neural activity from an exemplar subject supports a generalized decoding of the identity of a littermate. g Identity CCGP decoding with consecutive novel mouse interaction task data. h Representative image showing that IL^→NAcSh neural activity does not support a generalized decoding of the identity of novel mice. i Social interaction-dependent neural activities during littermate interactions showed significantly higher CCGP decoding performance than those from a null model, while those during novel conspecific interactions showed similar CCGP performance to those from a null model. Two-way ANOVA with Sidak’s multiple comparison test: n = 11, F_1,20 = 3.372, p = 0.0812, Littermate; Null-Actual: t₂₀ = 2.801, *p = 0.0220, Novel mouse; Null-Actual: t₂₀ = 0.2036, p = 0.9746. j Decoding familiarity using a CCGP analysis comparing social interaction-dependent IL^→NAcSh neural activity during littermate and novel mouse interactions. k A representative image of social interaction-dependent IL^→NAcSh neural activity from an exemplar subject in response to littermates and novel mice in the social familiarity CCGP analysis. l Social familiarity CCGP analysis comparing social interaction-dependent IL^→NAcSh neural activity during interactions with littermates or novel mice exhibited significantly higher decoding performance than that of null data. Two-tailed paired t-test: n = 9 mice, CCGP; t₈ = 6.500. ***p = 0.0002. Data presented as mean ± SEM; each dot represents one mouse. Statistics used one- or two-way RM ANOVA with two-sided Tukey-corrected post hoc comparisons. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. Source data are provided as a source data file. The mouse illustrations in (a) and (c) are reprinted from Cell Reports, Volume 35, Issue 6, Park, et al., Social isolation impairs the prefrontal-nucleus accumbens circuit subserving social recognition in mice, 109104, Copyright (2021), with permission from Elsevier.