Claims 61
From the paper Social isolation recruits amygdala–medial prefrontal cortex projections to escalate alcohol drinking in male mice (Nature Neuro) · claims file · checked 2026-10-21: ready
- fig1c-isolation-alcohol-sex-opposite
Social isolation changes alcohol licking in opposite directions in the two sexes, raising it in males and lowering it in females, a significant sex by housing interaction.
Social isolation produced opposite effects on alcohol licking in males and females ( N = 14 male and 11 female mice, two-way ANOVA, interaction: F (1, 23) = 21.04, *** P = 0.0001; Sidak’s multiple comparisons test), increasing alcohol licking in males while decreasing it in females.
Figure 1c legend; Results, Social isolation increases alcohol drinking in males and decreases it in females, paragraph 2How it will be judged
Per mouse, alcohol licks averaged over the last 3 baseline days and over the 11 isolation days (Methods, 2BC drinking). A two-way ANOVA of sex (between) by housing (within, repeated measures), as the error df of 23 with 25 mice implies; Sidak post hoc tests of housing within each sex. Reproduced when the interaction has p < 0.05, males rise and females fall, and the Sidak test is significant in males. Partially reproduced when the interaction holds but one sex's change is in the wrong direction.
- fig1-threeway-sex-liquid-housing
A three-way ANOVA over sex, liquid (alcohol, water) and housing finds a significant three-way interaction.
A three-way analysis of variance (ANOVA) across sex, liquid and housing condition revealed a significant interaction ( F (1, 46) = 14.68; P < 0.001), indicating that social isolation altered alcohol and water drinking differently across sexes.
Results, Social isolation increases alcohol drinking in males and decreases it in females, paragraph 2How it will be judged
The licks of fig1c-isolation-alcohol-sex-opposite, alcohol and water. The error df of 46 fits 50 mouse-by-liquid series less four groups, so the reading is housing as the repeated factor and sex and liquid as between factors; the fully repeated reading (liquid and housing within mouse) is shown as a variant. Reproduced when the three-way interaction has p < 0.05 under the first reading (paper p < 0.001).
- fig1d-male-escalation-time-course
In males alcohol drinking rises progressively over the days of isolation, peaking after about one week and staying high, with a significant effect of time.
Time course of escalated alcohol drinking during social isolation ( N = 14 male mice, one-way ANOVA, F (4, 65) = 2.83, * P = 0.03; post hoc Dunnett’s multiple comparisons test).
Figure 1d legend; Results, Social isolation increases alcohol drinking in males and decreases it in females, paragraph 2How it will be judged
The df give five time levels and an ordinary (not repeated) one-way ANOVA over 70 values. The levels are not stated; the readings are baseline (last 3 baseline days) and the 11 isolation days in four blocks of (a) days 1 to 3, 4 to 6, 7 to 9, 10 to 11, or (b) days 1 to 2, 3 to 5, 6 to 8, 9 to 11. Dunnett tests against baseline. Reproduced when under every reading the ANOVA has p < 0.05 and the largest block mean is at or after day 6; when the readings disagree the claim is methods_underspecified.
- fig1e-water-sex-not-housing
Isolation does not change water licking, while females lick more water than males (main effect of sex).
Social isolation did not impact water licking, but females drink more water than males ( N = 14 male and 11 female mice, two-way ANOVA, main effect of sex: F (1, 23) = 13.87, ** P = 0.001).
Figure 1e legend; Results, Social isolation increases alcohol drinking in males and decreases it in females, paragraph 2How it will be judged
Water licks as in fig1c-isolation-alcohol-sex-opposite, same ANOVA. The sex effect is reproduced when p < 0.05 with females higher. The absence of an isolation effect is judged on the same ANOVA at p < 0.05, for the housing main effect and the interaction; the size of the housing change is given either way. Both parts must hold for reproduced.
- ed1c-rank-alcohol-males
In group-housed males social rank correlates with alcohol licks, lower-ranked mice drinking more, and rank 4 mice drink more than the others.
Rank correlates with alcohol licks ( N = 26 male mice, Pearson correlation, r 2 = 0.25, Bonferroni adjusted * p = 0.01). Inset, Rank 4 mice drink more alcohol ( N = 26 male mice, one-way ANOVA, F(3, 22) = 4.68, * p = 0.01; Sidak’s post hoc, * p < 0.05).
Extended Data Fig. 1c legend; Results, Social isolation increases alcohol drinking in males and decreases it in females, paragraph 2How it will be judged
Rank under both M8 readings; licks are baseline alcohol licks (last 3 baseline days). The Bonferroni family is not stated; readings (a) two tests (alcohol and water, panels c and d), (b) no correction (the printed p = 0.01 matches the uncorrected p for r2 = 0.25 and 26 mice). Reproduced when r2 is within tolerance, licks rise toward subordinate ranks, the corrected p < 0.05 under every reading, and the one-way ANOVA over integer ranks has p < 0.05 with rank 4 highest.
- ed1d-rank-water-none
In group-housed males social rank does not correlate with water licks.
No correlation between social rank and water licks ( N = 26 male mice, Pearson correlation, r 2 = 0.06, p = 0.43).
Extended Data Fig. 1d legendHow it will be judged
A claim of absence on the paper's test, a Pearson correlation of rank (M8, both readings) with baseline water licks at p < 0.05. Reproduced when p >= 0.05; r2 is given either way.
- ed1e-rank-alcohol-isolation
During isolation rank still correlates with alcohol licks in males, but not with water licks.
Rank correlates with alcohol licks during isolation ( e ; N = 14 male mice, Pearson correlation, r 2 = 0.41, Bonferroni adjusted * p < 0.05), but not water ( f ; N = 14 male mice, Pearson correlation, r 2 = 0.002, Bonferroni adjusted p = 1).
Extended Data Fig. 1e,f legendHow it will be judged
Licks averaged over the 11 isolation days; rank under both M8 readings. The Bonferroni family is not stated; read as two tests (e and f). A four-test family (c to f) is not allowed: with r2 = 0.41 and 14 mice it gives p of about 0.054, against the printed p < 0.05. Reproduced when r2 for alcohol is within tolerance with corrected p < 0.05, and the water correlation has corrected p >= 0.05.
- ed1g-subordinate-licks-per-bout
Subordinate males increase their alcohol licks per bout during isolation (rank by isolation interaction), with no change in water licks per bout.
Subordinate males increase alcohol licks per bout during isolation ( g ; N = 4 male mice/group, two-way ANOVA, interaction: F (1, 6) = 20.84, *** p < 0.003, main effect of isolation: F(1, 6) = 28.84, ** p < 0.01; Sidak’s post hoc, *** p = 0.001 compared to group-housed), with no effect on water licks per bout ( h ).
Extended Data Fig. 1g,h legendHow it will be judged
Licks per bout (M1). Two groups of four, dominant and subordinate; which mice form them is not stated. Readings (a) rank 1 against the lowest rank of each cage, (b) rank 1 against rank 4 only. Two-way ANOVA, rank group by housing, housing repeated (df 6 = 8 mice less 2). Reproduced when the interaction has p < 0.05 with the subordinate rise significant (Sidak) and the water ANOVA shows no effect at p < 0.05, under every reading that yields four mice per group.
- ed1i-dominant-bout-number
Isolation increases the number of alcohol bouts in dominant males, with no change in water bouts.
Isolation increases alcohol bout number in dominant males ( i ; N = 4 male mice/group, two-way ANOVA, main effect of isolation: F (1, 6) = 10.23, * p = 0.01), with no effect on water bouts ( j ).
Extended Data Fig. 1i,j legendHow it will be judged
Bouts per session (M1), groups and ANOVA as in ed1g-subordinate-licks-per-bout. Reproduced when the main effect of isolation on alcohol bouts has p < 0.05 with dominants rising and no effect on water bouts at p < 0.05.
- ed1l-rank-alcohol-females
In females rank correlates with alcohol licks (one-tailed test, direction from males) but not with water licks.
Rank correlates with alcohol licks ( N = 9 female mice, one-tailed Pearson correlation, based on male mice, r 2 = 0.46, Bonferroni adjusted * p = 0.04), but not water licks ( m , Pearson correlation, r 2 = 0.08, Bonferroni adjusted p = 0.45).
Extended Data Fig. 1l,m legendHow it will be judged
Baseline licks; rank under both M8 readings; one-tailed in the male direction for alcohol. The Bonferroni family is not stated; read as the two tests (l, m), which matches the printed p = 0.04 for r2 = 0.46 and 9 mice. Reproduced when r2 is within tolerance, corrected one-tailed p < 0.05, and the water correlation has corrected p >= 0.05.
- ed1n-sex-alcohol-water
At baseline males lick more alcohol and less water than females (main effects of sex and of liquid).
Males show higher alcohol and lower water licking than females ( N = 26 male and 14 female mice/group, two-way ANOVA, main effect of sex: F(1, 38) = 10.77, ** p = 0.002; main effect of liquid: F(1, 38) = 17.29, *** p = 0.0002).
Extended Data Fig. 1n legendHow it will be judged
Baseline alcohol and water licks; two-way ANOVA sex (between) by liquid (within), df 38 = 40 mice less 2. Reproduced when both main effects have p < 0.05 and male alcohol exceeds female alcohol while male water is below female water. The paper is not consistent here: the Results say females show a "trend in higher baseline alcohol intake" (Results, Social isolation increases BLA–mPFC neuronal excitability, paragraph 2), so the direction of the alcohol difference is reported either way.
- ed1r-licks-blood-alcohol
Alcohol licks correlate strongly with blood alcohol concentration.
Alcohol licks (15% alcohol + 10% sucrose) correlates with blood alcohol concentrations, ( N = 19 male mice, Pearson’s correlation, r 2 = 0.67, **** p < 0.0001).
Extended Data Fig. 1r legend; Results, Social isolation increases alcohol drinking in males and decreases it in females, paragraph 2How it will be judged
Licks of the session after which trunk blood was taken (Methods, Blood alcohol concentration). Reproduced when r2 is within tolerance with p < 0.0001.
- fig2c-male-blampfc-excitability-up
After isolation BLA-mPFC neurons of males fire more action potentials across current injections than those of group-housed males.
Social isolation increases excitability of BLA–mPFC neurons in males (group-housed: n = 21 neurons/ N = 7 male mice; social isolation: n = 30 neurons/ N = 9 male mice, two-way ANOVA, interaction effect: F (25, 1225) = 4.3, **** P < 0.0001; Sidak’s multiple comparisons post hoc, * P < 0.05).
Figure 2c legend; Results, Social isolation increases BLA–mPFC neuronal excitability in males and decreases it in females, paragraph 2How it will be judged
M6, labeled neurons of males. Reproduced when the current by housing interaction has p < 0.05 (paper p < 0.0001), isolated neurons fire more at most steps, and at least one Sidak test across steps is significant, under both spike readings.
- fig2d-female-blampfc-excitability-down
After isolation BLA-mPFC neurons of females fire fewer action potentials across current injections than those of group-housed females.
Social isolation decreases excitability of BLA–mPFC neurons in females (group-housed: n = 18 neurons/ N = 9 female mice; social isolation: n = 15 neurons/ N = 7 female mice, two-way ANOVA, interaction effect: F (25, 775) = 2.16, *** P = 0.0009; Sidak’s multiple comparisons post hoc, * P < 0.05).
Figure 2d legend; Results, Social isolation increases BLA–mPFC neuronal excitability in males and decreases it in females, paragraph 3How it will be judged
M6, labeled neurons of females. Reproduced when the interaction has p < 0.05, isolated neurons fire less at most steps, and at least one Sidak test is significant, under both spike readings.
- fig2e-female-higher-baseline-excitability
Group-housed females have more excitable BLA-mPFC neurons than group-housed males.
Group-housed females exhibited higher baseline excitability of BLA–mPFC neurons compared with group-housed males (group-housed females: n = 18 neurons/ N = 9 female mice; group-housed males: n = 21 neurons/ N = 7 male mice, two-way ANOVA, interaction effect: F (25, 925) = 13.2, **** P < 0.0001; Sidak’s post hoc test, * P < 0.05).
Figure 2e legend; Results, Social isolation increases BLA–mPFC neuronal excitability in males and decreases it in females, paragraph 3How it will be judged
M6, group-housed labeled neurons. Reproduced when the current by sex interaction has p < 0.05, female neurons fire more at most steps, and at least one Sidak test is significant, under both spike readings.
- fig2-threeway-sex-current-housing
A three-way ANOVA over sex, current and housing finds a significant interaction for BLA-mPFC firing.
A three-way ANOVA across sex, current injection and housing condition revealed a significant interaction ( F (25, 2,000) = 5.6; P < 0.0001), indicating that social isolation altered BLA–mPFC excitability differently across sexes.
Results, Social isolation increases BLA–mPFC neuronal excitability in males and decreases it in females, paragraph 2How it will be judged
M6, the four groups of Fig. 2c,d; df 2000 = 25 times (84 neurons less 4 groups), current repeated. Reproduced when the three-way interaction has p < 0.05 under both spike readings.
- supt1-male-input-resistance-up
Isolation raises the input resistance of male BLA-mPFC neurons.
Post hoc two-way analyses showed that social isolation increased the intrinsic excitability of BLA–mPFC neurons in males, reflected by greater action potential firing across current injections (Fig. 2c ) and elevated input resistance (Supplementary Table 1 )
Results, Social isolation increases BLA–mPFC neuronal excitability in males and decreases it in females, paragraph 2; Supplementary Table 1How it will be judged
Input resistance from the steady-state response to the -120 pA step (Methods). The steady-state window is not given; readings the last 100 ms and the last 50 ms of the step. Unpaired t-test between housing groups, p < 0.05 (Supplementary Table 1 legend). Reproduced when isolated male neurons have higher input resistance with p < 0.05 under both readings. The table's values are not in the text read, so no number is checked.
- supt1-female-input-resistance-threshold
Isolation lowers the input resistance and raises the firing threshold of female BLA-mPFC neurons.
In contrast to males, isolation decreased BLA–mPFC excitability in females (Fig. 2d ), accompanied by reduced input resistance and a higher firing threshold (Supplementary Table 1 )
Results, Social isolation increases BLA–mPFC neuronal excitability in males and decreases it in females, paragraph 3; Supplementary Table 1How it will be judged
Input resistance as in supt1-male-input-resistance-up; threshold is the smallest current step eliciting an action potential (Methods), under both M6 spike readings. Unpaired t-tests, p < 0.05. Reproduced when both differ in the stated directions; partially when one does. No number is checked (table values not in the text read).
- ed2e-nonspecific-bla-excitability
Isolation also increases the excitability of nonspecific BLA neurons in males, a small but significant interaction, which the Results call minimal.
Social isolation increases excitability of non-specific BLA neurons in males (Group-housed: n = 50 neurons from N = 11 male mice; Social isolation: n = 54 neurons from N = 12 male mice; two-way ANOVA, interaction effect: F (25, 2550) = 1.58, * p = 0.03).
Extended Data Fig. 2e legend; Results, Social isolation increases BLA–mPFC neuronal excitability in males and decreases it in females, paragraph 2How it will be judged
M6, unlabeled neighbouring neurons of males. Reproduced when the interaction has p < 0.05 with isolated neurons firing more, under both spike readings. The paper gives no measure of "minimal"; the F and the size of the difference are reported beside those of fig2c-male-blampfc-excitability-up, not tested.
- ed2bcd-nonspecific-bla-no-rank
Nonspecific BLA neurons of males show no rank differences in excitability, group-housed or isolated, and isolation does not change their firing at 300 pA within any rank.
No detectable social rank-dependent differences were observed in non-specific BLA neurons from group-housed ( b ; Rank 1: n = 10 cells from N = 3 male mice; ... d . Social isolation did not alter the number of action potentials elicited by a 300 pA current injection across each social rank in males.
Extended Data Fig. 2b–d legend; Results, Social isolation increases BLA–mPFC neuronal excitability in males and decreases it in females, paragraph 3How it will be judged
A claim of absence. The legend names no test; the Statistics section gives two-way ANOVA with repeated measures for three or more groups, so b and c are tested by M6 with rank (1 to 3) in place of housing, and d by a two-way ANOVA of rank by housing on spikes at 300 pA, each at p < 0.05. Reproduced when no rank main effect or interaction reaches p < 0.05 in b and c, and no housing effect or interaction in d. The Results text cites these panels as Extended Data Fig. 3; the legends place them in Extended Data Fig. 2.
- ed3c-male-rank-blampfc-excitability
Among group-housed males, dominant (rank 1) mice have less excitable BLA-mPFC neurons than intermediate (rank 2) and subordinate (rank 3) mice.
Dominant male mice showed reduced basal BLA-mPFC excitability compared to Intermediate and Subordinate male mice (Rank 1: n = 13 neurons from N = 4 male mice; Rank 2: n = 9neurons from N = 4 male mice; Rank 3: n = 14 neurons from N = 5 male mice, two-way ANOVA, interaction effect: F (50, 825) = 4.25, **** p < 0.0001; Tukey’s post hoc, * p < 0.05, ** p < 0.01, compared to Rank 1).
Extended Data Fig. 3c legend; Results, Social isolation increases BLA–mPFC neuronal excitability in males and decreases it in females, paragraph 3How it will be judged
M6 with rank (1 to 3) as the group factor, group-housed labeled male neurons. Reproduced when the interaction has p < 0.05, rank 1 fires least, and Tukey tests show rank 1 below both rank 2 and rank 3 at some step; partially when only one of the two differs.
- ed3d-female-rank-blampfc-none
Group-housed females show no rank differences in BLA-mPFC excitability.
No detectable social rank-dependent differences were observed in BLA-mPFC neurons from group-housed female mice
Extended Data Fig. 3d legendHow it will be judged
A claim of absence. No test is named; M6 with rank as the group factor (the test of ed3c-male-rank-blampfc-excitability) at p < 0.05. Reproduced when neither the rank main effect nor the interaction reaches p < 0.05.
- ed4ef-chrimson-decreases-social
ChrimsonR stimulation of BLA-mPFC terminals decreases social interaction against tdTomato controls, both group-housed and after isolation.
BLA-mPFC stimulation decreased social interaction in ChrimsonR vs. tdTomato males under group-housed ( e ; two-way ANOVA, interaction: F(9,140) = 4.94, **** p < 0.0001; Sidak’s post hoc, *** p < 0.001) and social isolation ( f ; interaction: F(9,140) = 4.99, **** p < 0.0001; Sidak’s post hoc, *** p < 0.001) conditions ( N = 8 male mice/group).
Extended Data Fig. 4e,f legendHow it will be judged
Frames of the 3-min stimulation epoch (Methods, Optogenetic stimulation during resident-juvenile intruder task). The df imply 10 levels crossed with group, with 16 mice and no repeated-measures error (140 = 160 less 20). The levels are not named; readings (a) 10 UMAP behavioural clusters (legend c: unsupervised UMAP clustering of ChrimsonR and tdTomato frames), with the measure the fraction of frames per cluster; (b) 10 time bins of 18 s, with the measure the fraction of frames scored as interaction by the head-to-body distance of 60 pixels or less and angle of 135 degrees or less (Methods, SLEAP). UMAP parameters and the clustering of its embedding are not given; reading (a) uses the umap-learn defaults on the pairwise features the Methods list and the cluster count fixed at 10. Reproduced when, under every reading, the interaction has p < 0.05 in both housing conditions and ChrimsonR mice spend less time in interaction (Sidak significant for the interaction cluster or for some bins); methods_underspecified when readings disagree.
- ed4gm-locomotion-unchanged
Neither ChrimsonR stimulation nor NpHR inhibition changes total distance travelled against controls.
Total distance traveled did not differ between ChrimsonR and tdTomato mice (ChrimsonR: N = 15; tdTomato: N = 16 male mice).
Extended Data Fig. 4g,m legendHow it will be judged
A claim of absence. No test is named; the Statistics section gives t-tests for two groups, read as two-sided unpaired, p < 0.05. How distance was tracked is not given: behaviour was recorded with EthoVision (body-centre tracking) and pose was tracked with SLEAP; readings the EthoVision body centre where the dataset holds it, and the SLEAP skull base and haunch points. The epoch is not given; readings the 3-min stimulation epoch and the whole 6-min interaction. Reproduced when neither comparison (ChrimsonR 15 against tdTomato 16; NpHR 16 against eYFP 16) has p < 0.05 under any reading.
- ed4kl-nphr-social
NpHR inhibition of BLA-mPFC terminals decreases social avoidance when group-housed and increases social interaction after isolation.
BLA-mPFC inhibition decreased social avoidance in NpHR vs. eYFP males under group-housed ( k ; two-way ANOVA, interaction: F(6,98) = 4.02, ** p = 0.001; Sidak’s post hoc, *** p < 0.001) and increased social interaction under social isolation ( l ; two-way ANOVA, interaction: F(6,97) = 3.57, ** p = 0.003; Sidak’s post hoc, ** p < 0.01) conditions ( N = 8 male mice/group).
Extended Data Fig. 4k,l legend; Results, Social isolation increases BLA–mPFC neuronal excitability in males and decreases it in females, paragraph 4How it will be judged
As in ed4ef-chrimson-decreases-social with 7 levels (UMAP clusters of NpHR and eYFP frames, or 7 time bins). Reproduced when, under every reading, both interactions have p < 0.05, NpHR mice show less avoidance when group-housed and more interaction after isolation. The Results describe only the isolation effect ("increased social interaction selectively during isolation"); both parts are reported.
- ed5c-grin-no-effect-licks
GRIN lens implants do not change baseline alcohol licks.
GRIN lens implants do not alter baseline alcohol licks (No GRIN: N = 12 male mice; BLA GRIN: N = 22 male mice; mPFC GRIN: N = 6 male mice, one-way ANOVA, F (2, 37) = 0.47, p = 0.62).
Extended Data Fig. 5c legendHow it will be judged
A claim of absence on the paper's test, a one-way ANOVA over the three groups at p < 0.05. Reproduced when p >= 0.05.
- ed5d-cue-response-modest
Both BLA populations respond to the cue with a modest, transient increase below 0.5 z-score.
Both populations exhibited modest, transient increases in activity (<0.5 z-score) following cue onset
Extended Data Fig. 5d legendHow it will be judged
Cue-onset z-scores per M3 with cue onset in place of bout onset. The Methods give baselines only for bouts ("3- or 5-s"); readings 3 s and 5 s before cue onset (the intertrial interval of 53 to 73 s allows both). Reproduced, under both readings, when the population mean trace of each population peaks above 0 and below 0.5 within 3 s of cue onset and falls back toward 0 by the end of the 3-s cue.
- fig3f-blampfc-distinct-clusters
Hierarchical clustering of alcohol responses gives several functional clusters, and BLA-mPFC neurons distribute across them differently from nonspecific BLA neurons.
The distribution of BLA–mPFC neurons across functional clusters suggests that projection-defined BLA–mPFC neurons may represent a distinct functional subpopulation with specific response properties to alcohol, distinct from nonspecific BLA neurons (Fig. 3f ).
Figure 3d–f legends; Results, BLA–mPFC activity robustly encodes and predicts alcohol drinking in males, paragraph 3How it will be judged
One clustering of the 736 neurons of both populations pooled ("Hierarchical clustering of responses to alcohol for all recorded BLA neurons", Results, paragraph 3). Ward linkage, Euclidean distance on each neuron's mean alcohol z trace (M3), the cut set to the highest linkage giving at least one nonresponsive cluster (Methods, Agglomerative hierarchical clustering); "nonresponsive" read as a cluster whose mean trace stays within +-1.98 after onset (M4). The trace span is not given; readings -3 to +5 s and 0 to +3 s around onset. Percentage enrichment per cluster as in Methods, Enrichment analysis. The paper gives no test; the Statistics section names chi-square for distributions, so a chi-square test of population by cluster at p < 0.05. Reproduced when the test has p < 0.05 under both trace readings.
- ed5ef-alcohol-water-nonoverlapping
Neurons excited by alcohol and those excited by water are largely separate ensembles.
Nonspecific BLA and BLA–mPFC neurons excited to alcohol and water consisted of largely nonoverlapping neurons (Extended Data Fig. 5e,f ).
Extended Data Fig. 5e,f legend; Results, BLA–mPFC activity robustly encodes and predicts alcohol drinking in males, paragraph 2How it will be judged
Excited neurons per M4 for alcohol and for water bouts, in each population; the Wilcoxon reading of M4 is shown too. "Largely" is not quantified; read as fewer than half of the smaller excited set also excited by the other liquid. Reproduced when this holds in both populations under the threshold reading.
- fig3h-nonspecific-decoding-85
Nonspecific BLA population activity distinguishes alcohol from water drinking with about 85% accuracy.
We found that population activity in nonspecific BLA neurons could distinguish alcohol from water consumption with ~85% accuracy (Fig. 3h ).
Figure 3h; Results, BLA–mPFC activity robustly encodes and predicts alcohol drinking in males, paragraph 4How it will be judged
M5 on the 540 nonspecific BLA neurons of 10 mice. Reproduced when the mean accuracy over folds is within tolerance under every M5 reading; methods_underspecified when readings disagree.
- fig3h-blampfc-decodes-better
BLA-mPFC population activity decodes alcohol against water drinking more accurately than nonspecific BLA activity.
BLA–mPFC activity predicts alcohol drinking with greater accuracy than nonspecific BLA neurons (nonspecific BLA: n = 540 neurons/ N = 10; BLA–mPFC: n = 196 neurons/ N = 11 male mice, Kruskal–Wallis test, H (2) = 8.06, ** P = 0.009; Dunn’s multiple comparisons test, * P = 0.02).
Figure 3h legend; Results, BLA–mPFC activity robustly encodes and predicts alcohol drinking in males, paragraph 4How it will be judged
M5 for nonspecific BLA, BLA-mPFC and all BLA neurons (the three groups of the H(2) df, Extended Data Fig. 6b). Fold accuracies are the units. Reproduced when, under every M5 reading, the Kruskal-Wallis test has p < 0.05 and Dunn's test shows BLA-mPFC above nonspecific BLA at p < 0.05. The comparison of populations of different size (196 against 540 neurons) is as the paper did it; a size-matched subsample of nonspecific neurons is a variant only.
- ed6b-decoding-above-chance
Nonspecific BLA, BLA-mPFC and all BLA populations each decode alcohol against water above chance (shuffled data).
Population-level activity of non-specific BLA, BLA-mPFC, and all BLA neurons decodes alcohol vs. water drinking above chance (shuffled data), using a support vector machine ... two-way ANOVA, main effect: F(1,24) = 174.2, **** p < 0.0001; Sidak’s post hoc, **** p < 0.0001).
Extended Data Fig. 6b legendHow it will be judged
M5 with the shuffled control under both chance readings; two-way ANOVA population by data (real, shuffled) over fold accuracies (df 24 = 30 less 6). Reproduced when the main effect has p < 0.05 and every Sidak test of real against shuffled within a population has p < 0.05, under every M5 reading.
- fig3k-blampfc-response-predicts-bouts
Across mice, the mean BLA-mPFC response to alcohol correlates positively with the number of alcohol bouts.
We correlated the number of alcohol bouts with the mean BLA–mPFC response to alcohol and found a significant positive relationship, such that greater BLA–mPFC activity predicted higher alcohol consumption (Fig. 3k ).
Figure 3k legend; Results, BLA–mPFC activity robustly encodes and predicts alcohol drinking in males, paragraph 5How it will be judged
Per mouse, the mean over its BLA-mPFC neurons of the M3 alcohol response, against its number of alcohol bouts (M1) in the imaging session (mean over sessions if a mouse has more than one). Bonferroni family readings (a) four tests (Fig. 3i to l), (b) two tests (k and l). Reproduced when r2 is within tolerance, r is positive and the corrected two-sided p < 0.05 under both readings.
- fig3ijl-no-other-correlations
Nonspecific BLA responses do not correlate with alcohol or water bouts, and BLA-mPFC water responses do not correlate with water bouts.
No such relationship was observed for nonspecific BLA neurons, water consumption or cue responses (Fig. 3i,j,l and Extended Data Fig. 6e–j ).
Figure 3i,j,l legends; Extended Data Fig. 6e,f legend; Results, BLA–mPFC activity robustly encodes and predicts alcohol drinking in males, paragraph 5How it will be judged
A claim of absence, as fig3k-blampfc-response-predicts-bouts, on the paper's test (Pearson, Bonferroni over Fig. 3i to l) at p < 0.05. Reproduced when none of the nonspecific alcohol, nonspecific water and BLA-mPFC water correlations reaches corrected p < 0.05; r2 given for each.
- ed6ij-cue-responses-no-correlation
Cue-evoked responses of neither population correlate with alcohol bouts.
Scatter plots showing non-specific BLA neurons ( i , r 2 = 0.04, p = 0.53) and BLA-mPFC neurons ( j , r 2 = 0.005, p = 0.84) cue-evoked responses did not correlate with alcohol bouts.
Extended Data Fig. 6i,j legendHow it will be judged
A claim of absence on the paper's test (Pearson, p < 0.05). Per mouse mean cue response as in ed5d-cue-response-modest (mean z 0 to 3 s after cue onset) against alcohol bouts. Reproduced when neither correlation has p < 0.05.
- ed6k-bout-duration-not-number
Average bout duration does not correlate with the number of bouts.
No correlation was observed between the average bout duration and the number of bouts ( N = 11 male mice, Pearson correlation, r 2 = 0.0, p = 0.84).
Extended Data Fig. 6k legend; Results, BLA–mPFC activity robustly encodes and predicts alcohol drinking in males, paragraph 5How it will be judged
A claim of absence on the paper's test (Pearson, p < 0.05), alcohol bouts (M1) of the 11 BLA-mPFC mice. Reproduced when p >= 0.05.
- ed6cd-rank-stratified-correlation
The BLA-mPFC alcohol response correlates with alcohol bouts in lower-ranked (ranks 3 and 4) males but not in higher-ranked (ranks 1 and 2) males.
BLA-mPFC alcohol responses correlate with alcohol bout number in lower-ranked (Ranks 3-4) males ( d ; Pearson correlation, r 2 = 0.94, Bonferroni adjusted * p = 0.01) but not higher-ranked (Ranks 1-2) males ( c ; r 2 = 0.42, Bonferroni adjusted p = 1).
Extended Data Fig. 6c,d legend; Results, BLA–mPFC activity robustly encodes and predicts alcohol drinking in males, paragraph 5How it will be judged
The measures of fig3k-blampfc-response-predicts-bouts split by rank. Bonferroni readings (a) four tests (Extended Data Fig. 6c, d, g, h), (b) two tests (c and d). Reproduced when the lower-rank correlation is positive with corrected p < 0.05 and the higher-rank one has corrected p >= 0.05 under both readings, and both r2 are within tolerance.
- fig4e-chr2-increases-alcohol
Lick-triggered BLA-mPFC terminal stimulation increases alcohol licks in ChR2 mice against eYFP controls, returning to baseline in the recovery session.
BLA–mPFC terminal stimulation increased alcohol drinking in ChR2-expressing males compared with eYFP controls ( N = 6 male mice per group, two-way ANOVA, F (2, 20) = 4.7, * P = 0.02; Tukey’s multiple comparisons test).
Figure 4e legend; Results, BLA–mPFC terminal stimulation increases alcohol drinking in males, paragraph 1How it will be judged
Alcohol licks per session for baseline, stimulation and recovery sessions (Methods, Optogenetic stimulation during 2BC drinking). Two-way ANOVA group by session, session repeated (df 20 = 2 times 10). The legend does not say which effect F(2, 20) is; both the session main effect and the group by session interaction have these df, so both are readings. Baseline is not defined; readings the last baseline session and the mean of the last 3 baseline days (Methods, 2BC drinking). Reproduced when, under every reading, the effect has p < 0.05, Tukey shows ChR2 stimulation above ChR2 baseline, and ChR2 recovery is not above baseline at p < 0.05; methods_underspecified when readings disagree.
- fig4cd-chr2-no-sucrose-water-effect
BLA-mPFC terminal stimulation does not change sucrose or water drinking.
BLA–mPFC terminal stimulation did not alter sucrose ( c , N = 6 male mice per group) or water ( d , N = 6 male mice per group) drinking.
Figure 4c,d legends; Results, BLA–mPFC terminal stimulation increases alcohol drinking in males, paragraph 1How it will be judged
A claim of absence. The test is not named; the same two-way ANOVA as fig4e-chr2-increases-alcohol (group by session) on sucrose licks (1% sucrose baseline against stimulation session) and on water licks, at p < 0.05. Reproduced when neither has a significant interaction or group effect.
- fig4hi-chr2-longer-bouts
BLA-mPFC stimulation lengthens alcohol bouts and increases licks per bout in ChR2 mice.
BLA–mPFC stimulation increased alcohol bout duration ( h ; N = 6 male mice, two-sided unpaired t -test, **** P < 0.0001) and the number of alcohol licks per bout ( i ; N = 6 male mice, two-sided unpaired t -test, **** P < 0.0001).
Figure 4h,i legends; Results, BLA–mPFC terminal stimulation increases alcohol drinking in males, paragraph 2How it will be judged
Bouts (M1) of ChR2 mice, OFF against ON (stimulation session); an unpaired test with P < 0.0001 from 6 mice implies bouts as units. Every lick onset triggers stimulation in the test session, so OFF is another session, not named; readings the preceding baseline session and the recovery session. Reproduced when, under both readings, both measures are larger ON with two-sided unpaired p < 0.0001; partially when p < 0.05. The same test in eYFP mice (Fig. 4f) is reported; the paper says the effects were absent there.
- fig4g-chr2-bout-number-unchanged
Stimulation does not change the total number of alcohol bouts.
Analysis of lick microstructure revealed that BLA–mPFC activation prolonged individual alcohol drinking bouts (Fig. 4h ) and increased the number of licks per bout (Fig. 4i ), without changing the total number of bouts (Fig. 4g ).
Figure 4g inset; Results, BLA–mPFC terminal stimulation increases alcohol drinking in males, paragraph 2How it will be judged
A claim of absence; no test named. Read as a two-sided paired t-test over the 6 ChR2 mice, bouts OFF against ON (two groups per Statistics section), p < 0.05. Reproduced when p >= 0.05.
- fig5b-isolation-increases-mpfc-alcohol
Isolation increases mPFC neuron responses to alcohol.
Social isolation increased mPFC responses to alcohol (group-housed: n = 240 neurons/ N = 6 male mice; social isolation: n = 281 neurons/ N = 5 male mice, one-sided unpaired t -test, ** P = 0.006).
Figure 5b legend; Extended Data Fig. 8l legend; Results, Social isolation increases mPFC responses to alcohol while decreasing responses to sucrose in males, paragraph 2How it will be judged
M3 alcohol responses of mPFC neurons in the alcohol stimulation session before and after isolation (imaging was obtained during the stimulation sessions, Methods, Optogenetic stimulation and calcium imaging during cued-2BC drinking; the neuron counts equal those of Fig. 6c,f). Which trials enter is not said; readings (a) nonstimulated trials only, (b) all trials. Neurons are units. Reproduced when, under both readings, the isolated mean exceeds the group-housed mean with one-sided p < 0.05; the two-sided p of Extended Data Fig. 8l is reported (its p = 0.03 is not twice 0.006, so the two panels may use different trials).
- fig5c-isolation-no-water-change
Isolation does not change mPFC responses to water.
No detectable effect of social isolation was observed for mPFC responses to water (group-housed: n = 240 neurons/ N = 6 male mice; social isolation: n = 281 neurons/ N = 5 male mice).
Figure 5c legend; Results, Social isolation increases mPFC responses to alcohol while decreasing responses to sucrose in males, paragraph 2How it will be judged
A claim of absence on the test of fig5b-isolation-increases-mpfc-alcohol applied to water bouts, both trial readings; neither direction is stated, so two-sided unpaired t-test at p < 0.05. Reproduced when p >= 0.05 under both.
- fig5d-isolation-decreases-mpfc-sucrose
Isolation decreases mPFC neuron responses to sucrose.
In contrast to alcohol, social isolation decreased mPFC responses to sucrose (group-housed: n = 284 neurons/ N = 7 male mice; social isolation: n = 343 neurons/ N = 7 male mice, one-sided unpaired t -test, * P = 0.01).
Figure 5d legend; Results, Social isolation increases mPFC responses to alcohol while decreasing responses to sucrose in males, paragraph 2How it will be judged
As fig5b-isolation-increases-mpfc-alcohol for sucrose bouts of the sucrose stimulation sessions, both trial readings. Reproduced when, under both, the isolated mean is below the group-housed mean with one-sided p < 0.05. Extended Data Fig. 8g says the sucrose responses before and during isolation were not significantly different; that conflict is reported, and the two-sided p is given.
- ed7-mpfc-responses-stable
In group-housed mice, mPFC responses to alcohol, sucrose and social interaction do not change between two imaging time points 2 to 3 weeks apart.
No effect of time on mean mPFC responses to alcohol drinking in the cued 2-bottle choice drinking task (T1: n = 275 neurons and T2: n = 275 neurons from N = 4 male mice).
Extended Data Fig. 7b–d legend; Results, Social isolation increases mPFC responses to alcohol while decreasing responses to sucrose in males, paragraph 2How it will be judged
A claim of absence. No test is named; read as the two-sided unpaired t-test over neurons of fig5b-isolation-increases-mpfc-alcohol (the paired test is a variant for b, where T1 and T2 counts are equal), at p < 0.05, M3 responses, social bouts with the M3 social baselines. Reproduced when none of alcohol, sucrose and social interaction differs at p < 0.05.
- fig6c-stim-increases-mpfc-alcohol-gh
In group-housed mice, BLA-mPFC stimulation increases mPFC responses to alcohol, mimicking isolation.
BLA–mPFC stimulation increased mPFC alcohol responses in group-housed mice, mimicking social isolation ( n = 240 neurons/ N = 6 male mice, one-sided paired t -test, * P = 0.03).
Figure 6c legend; Results, BLA–mPFC stimulation mimics social-isolation-induced mPFC responses to alcohol and sucrose in males, paragraph 2How it will be judged
M3 alcohol responses of each neuron on stimulated and nonstimulated choice trials of the group-housed alcohol stimulation session; paired over neurons. Reproduced when stimulated exceeds nonstimulated with one-sided p < 0.05. Extended Data Fig. 8h says stimulation did not alter alcohol responses before isolation; that conflict is reported with the two-sided p.
- fig6de-stim-fewer-alcohol-inhibited
BLA-mPFC stimulation lowers the proportion of mPFC neurons inhibited by alcohol, without changing the proportion excited.
BLA–mPFC stimulation decreased inhibited mPFC alcohol responses ( e ; N = 6 male mice, one-sided paired t -test, t (5) = 4.6, ** P = 0.005) without affecting excited responses ( d ); threshold of ±1.98.
Figure 6d,e legend; Results, BLA–mPFC stimulation mimics social-isolation-induced mPFC responses to alcohol and sucrose in males, paragraph 2How it will be judged
Per mouse, the percentage of its neurons inhibited (and excited) per M4 threshold on stimulated and on nonstimulated trials, group-housed; paired over mice. Reproduced when the inhibited proportion falls with p < 0.05 and the excited proportion does not change at p < 0.05. The printed P = 0.005 matches a two-sided p for t(5) = 4.6 (one-sided would be about 0.003), so both sidedness readings are shown; the outcome must agree under both.
- fig6f-stim-decreases-mpfc-alcohol-si
After isolation, BLA-mPFC stimulation decreases mPFC responses to alcohol.
BLA–mPFC stimulation decreased mPFC alcohol responses during social isolation ( n = 281 neurons/ N = 5 male mice, one-sided paired t -test, * P = 0.02).
Figure 6f legend; Results, BLA–mPFC stimulation mimics social-isolation-induced mPFC responses to alcohol and sucrose in males, paragraph 2How it will be judged
As fig6c-stim-increases-mpfc-alcohol-gh in the isolation session. Reproduced when stimulated is below nonstimulated with one-sided p < 0.05. Extended Data Fig. 8j says stimulation did not alter alcohol responses during isolation; the conflict is reported with the two-sided p.
- fig6gj-stim-no-water-effect
BLA-mPFC stimulation does not change mPFC responses to water, nor the proportions of excited or inhibited water neurons, before or after isolation.
BLA–mPFC stimulation had no effect on mPFC water responses in group-housed ( g – i , N = 6) or social isolation ( j , N = 5) male mice.
Figure 6g–j legends; Extended Data Fig. 8q,r legend; Results, BLA–mPFC stimulation mimics social-isolation-induced mPFC responses to alcohol and sucrose in males, paragraph 2How it will be judged
A claim of absence on the tests of fig6c-stim-increases-mpfc-alcohol-gh (paired over neurons) and fig6de-stim-fewer-alcohol-inhibited (paired over mice) on water bouts, both housing conditions, at p < 0.05; sidedness not given for absence, so two-sided. Reproduced when none reaches p < 0.05.
- fig6k-stim-decreases-mpfc-sucrose-gh
In group-housed mice, BLA-mPFC stimulation decreases mPFC responses to sucrose, mimicking isolation.
BLA–mPFC stimulation decreased mPFC sucrose responses in group-housed mice, mimicking social isolation ( n = 284 neurons/ N = 7 male mice, one-sided paired t -test, ** P = 0.008).
Figure 6k legend; Extended Data Fig. 8c legend; Results, BLA–mPFC stimulation mimics social-isolation-induced mPFC responses to alcohol and sucrose in males, paragraph 3How it will be judged
As fig6c-stim-increases-mpfc-alcohol-gh on sucrose bouts of the group-housed sucrose stimulation session. Reproduced when stimulated is below nonstimulated with one-sided p < 0.05.
- fig6lm-stim-fewer-sucrose-excited
BLA-mPFC stimulation lowers the proportion of mPFC neurons excited by sucrose, without changing the proportion inhibited.
BLA–mPFC stimulation decreased excited mPFC sucrose responses ( l ; N = 7 male mice, one-sided paired t -test, t (6) = 2.76, * P = 0.03) without affecting inhibited responses ( m ); threshold of ±1.98.
Figure 6l,m legend; Results, BLA–mPFC stimulation mimics social-isolation-induced mPFC responses to alcohol and sucrose in males, paragraph 3How it will be judged
As fig6de-stim-fewer-alcohol-inhibited on sucrose, group-housed. Reproduced when the excited proportion falls with p < 0.05 and the inhibited proportion does not change at p < 0.05. The printed P = 0.03 matches a two-sided p for t(6) = 2.76; both sidedness readings are shown.
- fig6n-isolation-occludes-sucrose-effect
After isolation, BLA-mPFC stimulation no longer changes mPFC responses to sucrose.
Social isolation occluded BLA–mPFC stimulation effects on mPFC sucrose responses ( n = 343 neurons/ N = 7 male mice).
Figure 6n legend; Extended Data Fig. 8e legend; Results, BLA–mPFC stimulation mimics social-isolation-induced mPFC responses to alcohol and sucrose in males, paragraph 3How it will be judged
A claim of absence on the paper's paired t-test over neurons (as fig6k), stimulated against nonstimulated in the isolation sucrose session; read two-sided at p < 0.05. Reproduced when p >= 0.05.
- ed8n-stim-fewer-sucrose-inhibited-si
During isolation, BLA-mPFC stimulation lowers the proportion of mPFC neurons inhibited by sucrose, with no change in those excited.
BLA-mPFC stimulation significantly decreased the proportion of mPFC neurons inhibited to sucrose ( n ; paired t-test, ** p = 0.005) with no effect on mPFC neurons excited to sucrose ( m ), based on a mean Z-score response ±1.98, during social isolation.
Extended Data Fig. 8m,n legendHow it will be judged
As fig6lm-stim-fewer-sucrose-excited in the isolation sucrose session, paired over mice, two-sided. Reproduced when the inhibited proportion falls with p < 0.05 and the excited proportion does not change at p < 0.05.
- ed8op-stim-no-alcohol-proportions-si
During isolation, BLA-mPFC stimulation does not change the proportions of mPFC neurons excited or inhibited by alcohol.
BLA-mPFC stimulation did not alter the proportion of mPFC neurons responsive to alcohol during social isolation.
Extended Data Fig. 8o,p legendHow it will be judged
A claim of absence. No test is named; the test of the neighbouring panels (Extended Data Fig. 8m,n), a two-sided paired t-test over the 5 mice of the isolation alcohol stimulation session, on the percentages of neurons excited and inhibited per M4 threshold, stimulated against nonstimulated trials, at p < 0.05. Reproduced when neither reaches p < 0.05.
- fig6q-nphr-decreases-alcohol-bouts
During isolation, NpHR inhibition of BLA-mPFC terminals decreases the number of alcohol bouts, without changing water drinking.
BLA–mPFC inhibition decreased alcohol bouts ( N = 8 male mice per group, two-way ANOVA, main effect of stimulation: F (1,13) = 9.707, ** P = 0.008; Sidak’s post hoc test, * P = 0.05).
Figure 6q legend; Results, BLA–mPFC inhibition during social isolation reduces alcohol drinking in males, paragraph 1How it will be judged
Alcohol bouts (M1) on inhibited and noninhibited choice trials of the isolation inhibition session; two-way ANOVA group by stimulation, stimulation repeated (df 13 implies 15 mice). Bouts read as counts per condition; per-trial rates are a variant (equal halves of choice trials make them proportional). Reproduced when the stimulation effect has p < 0.05 with fewer bouts on inhibited trials in NpHR mice (Sidak p <= 0.05), and the same ANOVA on water bouts finds no stimulation effect or interaction at p < 0.05.
- ed9c-coregistered-fraction
About 30% of mPFC neurons are co-registered across all four social sessions.
~ 30% of mPFC neurons co-registered across all four sessions
Extended Data Fig. 9c legendHow it will be judged
CellReg on CNMF-E footprints (Methods). The denominator is not given; readings (a) neurons of the reference session, (b) the mean neuron count over the four sessions, (c) all distinct registered cells. Reproduced when one reading is within tolerance, as the protocol allows for denominators. If the dataset holds only the co-registered 199 neurons the claim cannot be tested.
- ed9e-isolation-decreases-social-response
Isolation decreases mPFC responses to social interaction.
Isolation decreases mPFC responses to interaction ( n = 199 neurons/ N = 6 male mice, two-way RM-ANOVA, interaction: F (49, 19404) = 1.538, ** p = 0.009).
Extended Data Fig. 9e legendHow it will be judged
M3 z traces of the 199 co-registered neurons around interaction-bout onsets (UMAP social clusters, Methods, SLEAP) on no-stimulation sessions, before and after isolation. 50 time points (df 49) at 10 Hz; readings 0 to 5 s after onset, and -2.5 to 2.5 s. Baselines 3 s and 5 s (M3). Reproduced when the time by housing interaction has p < 0.05 and the isolation trace is below the group-housed trace after onset, under every reading.
- ed9fg-stim-social-response
BLA-mPFC stimulation does not change mPFC responses to social interaction before isolation but increases them during isolation.
BLA-mPFC stimulation did not affect mPFC responses before isolation ( n = 199 co-registered neurons from N = 6 male mice). g . BLA-mPFC stimulation increases mPFC responses during isolation ( n = 199 neurons/ N = 6 male mice, two-way ANOVA, interaction: F (49, 7350) = 1.631, ** p = 0.003).
Extended Data Fig. 9f,g legendHow it will be judged
As ed9e-isolation-decreases-social-response, stimulation against no-stimulation session in each housing condition. Reproduced when the isolation interaction has p < 0.05 with stimulated above and the group-housed interaction has p >= 0.05. The error df 7350 (150 times 49) does not match 199 neurons; this is reported, not resolved.
- ed9hi-stim-inverts-social-response
In group-housed mice, mPFC social responses with and without stimulation are negatively correlated across neurons (stimulation inverts their sign); during isolation they are uncorrelated.
Stimulation and no-stimulation mPFC responses are negatively correlated in group-housed mice ( h ; n = 199 neurons/ N = 6 male mice, Pearson correlation, r 2 = 0.02, Bonferroni adjusted * p = 0.04) but uncorrelated during isolation ( i ; r 2 = 0.0, p = 1).
Extended Data Fig. 9h,i legend; Results, BLA–mPFC stimulation mimics social-isolation-induced mPFC responses to alcohol and sucrose in males, paragraph 4How it will be judged
Per neuron, scalar M3 social response (0 to 3 s) with and without stimulation. The Bonferroni family is not stated. A two-test family cannot give the printed p = 0.04: with r2 = 0.02 and 199 neurons the uncorrected p is about 0.046, and r2 below 0.025 keeps the doubled p above 0.05. Readings (a) uncorrected, (b) two tests, reported both; the outcome is judged on (a), the only reading consistent with the printed numbers. Reproduced when the group-housed r is negative with p < 0.05 and the isolation correlation has p >= 0.05, under both baseline readings.
- ed9k-o-social-decoding
In group-housed mice mPFC activity decodes social interaction above chance with and without stimulation; isolation abolishes decoding without stimulation and lowers accuracy, and BLA-mPFC stimulation raises accuracy in both housing conditions, rescuing it after isolation.
mPFC decodes interaction above chance under both conditions (two-way ANOVA, main effect of group: F(1,36) = 54.29, **** p < 0.0001; ... m . Isolation decreases mPFC decoding accuracy (unpaired t-test, * p = 0.02). n,o . BLA-mPFC stimulation increases decoding accuracy in group-housed ( n , unpaired t-test, * p = 0.01) and isolation ( o , unpaired t-test, * p = 0.005) conditions.
Extended Data Fig. 9k–o legendsHow it will be judged
M5 with interaction against noninteraction bouts (noninteraction bouts over 20 s split, bouts 2 s apart; Methods) for each of the four sessions. The df 36 imply 10 accuracy values per cell, so k = 10 folds here (readings: 10 folds; 5 folds repeated twice). Panels k and l are read as one two-way ANOVA per housing condition, data (real, shuffled) by stimulation, over fold accuracies (2 by 2 by 10 values, error df 36). Reproduced when, under every reading, the group-housed ANOVA has a real against shuffled main effect at p < 0.05 with real above shuffled in both stimulation conditions (Sidak); the isolation ANOVA has a data by stimulation interaction at p < 0.05, with isolation no-stimulation accuracy not above shuffled and stimulation accuracy above shuffled (Sidak); isolation no-stimulation accuracy is below group-housed no-stimulation (unpaired p < 0.05); and stimulation exceeds no-stimulation in both housing conditions (unpaired p < 0.05). Partially reproduced when some but not all of these hold.
- ed9q-stim-recruits-isolation-ensemble
BLA-mPFC stimulation recruits an mPFC ensemble resembling the isolation pattern, with greater overlap of excited neurons than in the group-housed comparison.
BLA-mPFC stimulation recruits a mPFC ensemble resembling isolation-induced patterns, with greater overlap of excited neurons vs. group-housed ( n = 199 neurons/ N = 6 male mice , Chi 2 = 6.94, * p = 0.03).
Extended Data Fig. 9p–r legends; Results, BLA–mPFC stimulation mimics social-isolation-induced mPFC responses to alcohol and sucrose in males, paragraph 4How it will be judged
Excited neurons by M4 in each session. The contingency table is not given; readings (a) among neurons excited by stimulation in the group-housed session, those also excited in the isolation no-stimulation session, against those also excited in the group-housed no-stimulation session; (b) clusters from the concatenated clustering (Extended Data Fig. 9p) cross-tabulated by session. Reproduced when, under every reading, overlap with the isolation pattern is greater with chi-square p < 0.05; methods_underspecified when readings disagree.
Pipeline
- Paper
- done (2 of 2)
- Claims
- done
- Claims check
- done (ready)
- Attempt
- todo
- Questions
- none
- Verified
- none