Datasets / DANDI:000718

DANDI:000718

✓ Verified by an independent agent

Offline ensemble co-reactivation links memories across days

cell population imaging microscopy House mouse Converted by CatalystNeuro 40.6 GB 20 files
  • 2 reproduced
  • 2 not reproduced
  • 21 blocked

Counted separately: 1 check that the dataset is the data behind the paper.

  • 1 reproduced

View on DANDI Paper JSON How to read outcomes

Claims

From the paper Offline ensemble co-reactivation links memories across days

Reproducing Zaki et al. (2024) from dandiset 000718 6

Notebook reproduce_zaki_2024.ipynb

  • Not reproducedFigure 2d and 2e; Extended Data Figure 6e; Results, Co-reactivation occurs more during wake, paragraph 1

    Offline population bursts coincide with the mouse briefly slowing down about 1 s before the burst and resuming its movement about 1 to 2 s later.

    Mice showed a robust and brief slowing down around 1 s before each burst event, before increasing locomotion back up around 2 s later.

    Figure 2d and 2e; Extended Data Figure 6e; Results, Co-reactivation occurs more during wake, paragraph 1

    Telemetry movement in the second before a burst is not lower than 3 to 5 s before it in any of the three recordings under either reading of time zero (peak, one-sided Wilcoxon p = 0.76, 0.89, 0.39; onset, p = 0.88, 0.94, 0.41); 30 to 60% of bursts have lower movement before them. Movement 1 to 2 s after the peak is lower than in the second before it in two of three recordings (0.155 against 0.225, and 0.129 against 0.257). The telemetry activity signal is a coarse proxy whose relation to calcium activity is weak (movement decoded at r = 0.06 to 0.24 at zero shift; Spearman r with the population trace 0.07 to 0.32 in 1 s bins). The files describe it as motion of the probe relative to the receiver and a proxy for locomotion; its mean magnitude is 4 to 15 times higher in wake than in quiet wake epochs. The test is not calibrated, giving p = 0.016 at 300 random times in Ca_EEG2-1 (0.9999 and 0.92 in the others). The outcome says this proxy does not show the pause, and is weak evidence that the pause is absent.

    How the test differed from the paper’s method

    • Movement is the magnitude of the HD-X02 telemetry activity signal; the paper tracked locomotion with a webcam, whose video is not deposited.
    • Tested on the EEG/EMG experiment's offline recordings; the paper shows the effect for the single-experience cohort (Figure 2d and 2e) and the inhibitory tagging cohort (Extended Data Figure 6e).
  • Not reproducedFigure 5c (left); Methods, "Ensemble co-participation analysis"

    After a strong (high-shock) aversive experience, during wake in the offline period the overlap ensemble co-bursts more with the neutral ensemble than with the aversive ensemble (t4 = 4.94, P = 0.016).

    Left, wake high-shock mice had higher co-bursting of overlap × neutral than overlap × aversive

    Figure 5c (left); Methods, "Ensemble co-participation analysis"

    What the paper reports

    • t4 = 4.94, p = 0.016 (n = 5 high-shock mice)

    In Ca_EEG2-1's first offline recording after aversive encoding, under reading (a), wake and quiet wake together (all 87 bursts), OxN is below OxA (26.4 against 35.6% inclusive, 8.0 against 17.2% exclusive; 7 against 15 discordant bursts, McNemar p = 0.13). Reading (a) is the one the Methods imply. Their scoring (6 s epochs, two-state Gaussian mixture on EMG) applied to this mouse's log EMG RMS over the 12 h offline period puts 94.1% of quiet-wake frames and 98.5% of wake frames in wake (97.9% of quiet wake when fitted on the first 6 h), and all 87 bursts in wake. Under reading (b), wake alone, OxN exceeds OxA (32.0 against 28.0% of 25 bursts; 8.0 against 4.0%) on 2 against 1 discordant bursts (McNemar p = 1.0). Under the Methods' wake the order holds at thresholds 1.5 (34.9 against 41.1%) and 2.5 (21.7 against 30.0%), and no difference reaches p < 0.05. Outside bursts both are at most 1.02% of 0.5 s windows. The neutral, aversive and overlap ensembles each have more of their cells in bursts than the remaining cells (median differences 0.005 to 0.010, Wilcoxon p = 0.041 to 0.0003). The paper's statistic (t4 = 4.94, P = 0.016, across five mice) cannot be tested. The evidence against the paper is weak; one 10 min recording of 24 for one mouse of five, with ensemble traces z-scored within it, so levels are not comparable with the paper's source data for this mouse over all 24 recordings (10.37% OxN against 8.39% OxA in wake). Counted inclusively, co-participation here (26 to 36%) is 2.5 to 4 times those values; with the third ensemble excluded (8.0 and 17.2%) it is on their scale, with OxN the lower.

    How the test differed from the paper’s method

    • One high-shock mouse of five, and its first offline recording of 24 (10 min of 4 h).
    • The Methods z-score the ensemble traces over all 24 offline recordings concatenated; here only one is deposited, so they are z-scored within it.
    • The criterion fixed four readings (quiet wake in or out of wake; co-participation with or without the third ensemble) and would count the claim under any of them. After the first verification the outcome is judged on wake including quiet wake, as the Methods' own EMG scoring puts quiet wake in wake; under wake alone OxN exceeds OxA without a significant difference, which the criterion would have called partially reproduced.
    • The EMG feature of the Methods' scoring (log RMS per 6 s epoch) is chosen here, after the data were seen; the Methods do not name one. Without the log the mixture does not separate sleep from wake.
    • Bursts are assigned the sleep state of their peak frame, a choice made here.
  • ReproducedExtended Data Figure 9b

    In the calcium imaging and EEG/EMG experiment, high-shock mice froze more during aversive recall than low-shock mice (t7 = 8.99, p = 0.020).

    High-shock mice froze more during Aversive recall than low-shock mice

    Extended Data Figure 9b

    What the paper reports

    • t7 = 8.99, p = 0.02 (n = 4 low-shock and 5 high-shock mice)

    Recall1 freezing is 64.09% for the high-shock mouse and 8.82% for the low-shock mouse (difference 55.27 percentage points). The 30 s bins differ by Mann-Whitney (U = 100, p = 0.00018); 15 s and 60 s bins give p = 8.9e-07 and 0.0079. The paper's t-test (t7 = 8.99, p = 0.020) over 5 and 4 mice cannot be run with one mouse per group, so its statistic is not tested. The bins are consecutive parts of one session per mouse and not independent, so the p-values overstate the evidence.

    How the test differed from the paper’s method

    • One mouse per group (the paper has 5 high-shock and 4 low-shock mice); the test is within sessions, over 30 s bins, chosen here.
  • ReproducedExtended Data Figure 3n and 3o

    The mean population activity during offline periods is more right-skewed than in shuffled data, showing brief burst-like activations that need coordinated activity of groups of neurons.

    The true skew was greater than the shuffled skew in every mouse ... supporting the idea that the mean population activity undergoes brief burst-like activations requiring the coordinated activity of groups of neurons

    Extended Data Figure 3n and 3o

    What the paper reports

    • t7 = 13.36, p = 3 × 10⁻⁶ (n = 8 mice)

    The population trace's skewness is 1.65, 1.34 and 1.72 in the three offline recordings, above all 1000 circular-shift surrogates in each (surrogate means 0.60, 0.40 and 0.37, 95th percentiles 0.72, 0.51 and 0.48). Both arms are lower than in the paper's source data for Extended Data Figure 3o (another cohort, 1 h at 30 Hz); the true skews are below the paper's lowest (1.76 to 3.65) and two of three surrogate means are below its lowest shuffled value (0.47 to 1.91), so the chance level is not shown to match the paper's. The outcome rests on the true skew exceeding every surrogate. On synthetic data the test separates coordinated activity (skew 8.54 against a surrogate 95th percentile of 0.94) from independent cells (0.98 against 1.09).

    How the test differed from the paper’s method

    • Run on the EEG/EMG experiment's three 10 min offline recordings; the paper shows the test for another cohort (Extended Data Figure 3o), which is not deposited.
  • BlockedResults, Neutral ensemble recruited into population bursts, paragraph 1; Extended Data Figure 3l and 3m· Methods underspecifiedverifier disagreed at first

    During offline periods, hippocampal population bursts occur more often than expected from shuffled neuronal activity, so CA1 neurons fire in a more coordinated way than chance.

    True burst frequencies were greater than shuffled burst frequencies in every mouse ... suggesting that during the offline period, hippocampal CA1 neurons fire in a more coordinated manner than would be expected from shuffled neuronal activities

    Results, Neutral ensemble recruited into population bursts, paragraph 1; Extended Data Figure 3l and 3m

    What the paper reports

    • t7 = 6.159, p = 4.6 × 10⁻⁴ (n = 8 mice)

    The analysis ran. Under the criterion's procedure (surrogates z-scored and thresholded at z = 2 as the data) the true burst rate is below the surrogate mean in all three offline recordings (5.30, 6.40 and 8.71 per min against 8.48 to 10.94; true over surrogate mean 0.63 to 0.80), which would be not reproduced. The paper's chance level for this test (source data for Extended Data Figure 3m, another cohort) is 3.38 to 4.39 surrogate bursts per min against 4.51 to 4.92 true, surrogate over true 0.72 to 0.96; this procedure gives 1.26 to 1.60, far from it, so the procedures differ and the claim is blocked as the protocol requires. The Methods do not say how bursts are counted in a surrogate. Variants disagree; merging events closer than 1 s leaves the true rate below the surrogates (4.40 to 7.31 against 6.56 to 8.52 per min), and thresholding surrogates at the true trace's own mean + 2 SD puts the true rate far above them (5.30 to 8.71 against at most 0.50 per min). At z = 2.5 the true rate is above the surrogate mean in 2 of 3 recordings and above the 95th percentile in none. On synthetic data with 6 coordinated events per minute, the criterion's test gives 5.6 true bursts per min against a surrogate mean of 34.98, and with independent cells 34.8 against 37.36, so it cannot detect coordination; the fixed-threshold variant separates the two (5.6 against 0.00, 34.8 against 38.37), but its chance level in the data (surrogate means at most 0.10 per min) is far below the paper's.

    How the test differed from the paper’s method

    • Run on the three 10 min offline recordings of the EEG/EMG experiment at 15 Hz; the paper shows the test for the one-hour recordings at 30 Hz of the single-experience cohort (Extended Data Figure 3), which is not deposited.
    • Burst events taken as maximal runs of frames above threshold, as the Methods do not define an event further.

Not tested in a notebook 20

20 data missing from the dataset
  • BlockedFigure 1a to 1c; Results, Retrospective memory linking across days, paragraph 1· Data missing from the dataset

    Fear spreads from an aversive context to a neutral context experienced 2 days before (retrospective linking) but not to one experienced 2 days after (prospective), with no difference in aversive recall between the groups.

    There was no difference in freezing in the aversive context between groups ... In the retrospective group, mice froze more in the neutral context compared with in the novel context ... in the prospective group, there was no difference in freezing between the neutral and novel contexts

    Figure 1a to 1c; Results, Retrospective memory linking across days, paragraph 1

    What the paper reports

    • aversive recall freezing, prospective against retrospective (Figure 1b): t34 = 0.36, p = 0.72 (n = 16 retrospective and 20 prospective mice)
    • direction by context interaction (Figure 1c): F1,32 = 4.90, p = 0.034 (n = 8 retrospective neutral, 8 retrospective novel, 12 prospective neutral and 8 prospective novel mice)
    • neutral against novel, retrospective group (Figure 1c): post hoc test, t32 = 2.586, p = 0.029
    • neutral against novel, prospective group (Figure 1c): post hoc test, t32 = 0.452, p = 0.6546

    Behaviour-only experiment; its mice are not in the dandiset. The paper's source-data repository holds the per-mouse freezing plotted in Figure 1, but not the recordings they come from; these claims were not tested on figure values.

  • BlockedFigure 1d to 1f· Data missing from the dataset

    High-shock mice froze more than low-shock mice in the aversive context, and froze more in the neutral than in the novel context, while low-shock mice did not.

    Mice froze more in the aversive context after receiving a high shock versus low shock ... High-shock mice froze more in the neutral versus novel contexts ... and low-shock mice froze no differently

    Figure 1d to 1f

    What the paper reports

    • aversive recall freezing, high against low shock (Figure 1e): t18,8 = 5.877, p = 1.2 × 10⁻⁵ (n = 10 low-shock and 12 high-shock mice)
    • effect of context (Figure 1f): F1,20 = 17.32, p = 4.8 × 10⁻⁵ (n = 10 low-shock and 12 high-shock mice)
    • context by amplitude interaction (Figure 1f): F1,20 = 4.99, p = 0.037
    • neutral against novel, high-shock mice (Figure 1f): t11 = 4.37, p = 0.002
    • neutral against novel, low-shock mice (Figure 1f): t9 = 1.23, p = 0.249

    The one-hour imaging cohort of Figure 1d to 1f is not in the dandiset. Its EEG-cohort counterpart is tested in ed9b-aversive-freezing-high-vs-low.

  • BlockedFigure 1g· Data missing from the dataset

    In high-shock mice, aversive recall freezing correlates with the strength of retrospective memory linking (R2 = 0.45, P = 0.016), but not in low-shock mice.

    Aversive memory strength was correlated with the strength of retrospective memory linking in high-shock mice ... but not in low-shock mice

    Figure 1g

    What the paper reports

    • high-shock mice: R2 = 0.45, p = 0.016 (n = 12 high-shock mice)
    • low-shock mice: R2 = 0.0003, p = 0.963 (n = 10 low-shock mice)

    Needs the one-hour imaging cohort's freezing in three contexts; not in the dandiset.

  • BlockedExtended Data Figure 1l to 1p; Results, Retrospective memory linking across days, paragraph 3· Data missing from the dataset

    Chemogenetic inhibition of the hippocampus in the 12 h after aversive encoding prevents retrospective memory linking and leaves the aversive memory intact.

    We found that this prevented retrospective memory linking (that is, selective freezing in the neutral context compared with in the novel context) while leaving the aversive memory intact

    Extended Data Figure 1l to 1p; Results, Retrospective memory linking across days, paragraph 3

    What the paper reports

    • group by context interaction (Extended Data Figure 1n): F1,42 = 4.00, p = 0.05 (n = 12 saline neutral, 10 saline novel, 12 PSEM neutral and 12 PSEM novel mice)
    • neutral against novel, saline group (Extended Data Figure 1n): post hoc test, t19.84 = 2.57, p = 0.03
    • neutral against novel, PSEM group (Extended Data Figure 1n): post hoc test, t22 = 0.31, p = 0.76
    • aversive recall freezing, PSEM against saline (Extended Data Figure 1p): t13.9 = 0.32, p = 0.748 (n = 7 saline and 9 PSEM mice)

    The chemogenetic inhibition experiment is behaviour only and not in the dandiset.

  • BlockedExtended Data Figure 2g to 2i; Results, Retrospective memory linking across days, paragraph 2· Data missing from the dataset

    A cocaine experience also drives retrospective memory linking to a neutral context experienced 2 days before.

    we also found evidence that a highly salient, appetitive experience (that is, cocaine exposure) also drove retrospective memory linking to a neutral context memory formed 2 days before

    Extended Data Figure 2g to 2i; Results, Retrospective memory linking across days, paragraph 2

    What the paper reports

    • locomotion during neutral recall, cocaine against saline (Extended Data Figure 2i): t54 = 2.85, p = 0.01 (n = 28 cocaine and 28 saline mice)
    • locomotion during novel recall, cocaine against saline (Extended Data Figure 2i): t54 = 1.83, p = 0.07 (n = 28 cocaine and 28 saline mice)

    The cocaine experiment is behaviour only and not in the dandiset.

  • BlockedExtended Data Figure 3e to 3k; Results, Offline reactivation of a past neutral ensemble, paragraph 2· Data missing from the dataset

    Cells active during aversive encoding are reactivated offline and preferentially during aversive recall.

    we found that the ensemble of cells active during aversive encoding was reactivated offline and preferentially reactivated during aversive memory recall

    Extended Data Figure 3e to 3k; Results, Offline reactivation of a past neutral ensemble, paragraph 2

    What the paper reports

    • recall reactivation of offline-reactivated against other aversive encoding cells (Extended Data Figure 3f): t7 = 19.41, p = 2 × 10⁻⁷ (n = 8 mice)
    • fraction previously active in aversive encoding, aversive against novel recall (Extended Data Figure 3g): t7 = 6.897, p = 2 × 10⁻⁴ (n = 8 mice)
    • offline activity, aversive against remaining ensemble (Extended Data Figure 3h): t7 = 8.538, p = 6 × 10⁻⁵ (n = 8 mice)

    The single-experience imaging cohort is not in the dandiset.

  • BlockedExtended Data Figure 4f· Data missing from the dataset

    During offline 2 the neutral ensemble is less active than the aversive and overlap ensembles in low-shock mice but no differently active in high-shock mice.

    The Neutral ensemble was less active than the Aversive and Overlap ensembles ... In high-shock mice ... these three ensembles were no differently active from each other

    Extended Data Figure 4f

    What the paper reports

    • ensemble by amplitude interaction: F3,331 = 5.36, p = 0.0013 (n = 7 low-shock and 8 high-shock mice)
    • neutral against aversive and overlap, low-shock mice: t12 = 2.83, p = 0.03
    • neutral, aversive and overlap, high-shock mice: F2,14 = 1.52, p = 0.25

    The one-hour offline cohort is not in the dandiset. The deposited high-shock offline recording gives one mouse of the EEG cohort, for which the paper does not report this analysis.

  • BlockedFigure 3c· Data missing from the dataset

    During offline 1 a greater percentage of the neutral ensemble participates in bursts than of the remaining ensemble, in both shock groups.

    During offline 1 in the low- and high-shock groups, a greater percentage of the neutral ensemble participated in bursts than the remaining ensemble

    Figure 3c

    What the paper reports

    • ensemble: F1,13 = 16.33, p = 0.001 (n = 7 low-shock and 8 high-shock mice)
    • ensemble by amplitude interaction: F1,13 = 0.0058, p = 0.94

    One-hour offline cohort not deposited. In the EEG cohort, the offline-day-1 recording of Ca_EEG3-4 is deposited without registration, and Ca_EEG2-1 has none.

  • BlockedFigure 3d to 3f· Data missing from the dataset

    During offline 2 the neutral ensemble preferentially participates in bursts relative to the remaining ensemble in high-shock mice (t7 = 4.88) but not in low-shock mice.

    High-shock mice showed higher neutral ensemble participation relative to the remaining ensemble ... low-shock mice showed no difference in ensemble participation

    Figure 3d to 3f

    What the paper reports

    • ensemble by amplitude interaction: F1,13 = 5.186, p = 0.04 (n = 7 low-shock and 8 high-shock mice)
    • neutral against remaining, high-shock mice: t7 = 4.88, p = 0.0036
    • neutral against remaining, low-shock mice: t6 = 1.33, p = 0.23

    One-hour offline cohort not deposited. The notebook shows, as context only, that in Ca_EEG2-1's offline recording the neutral ensemble has more of its cells in bursts than the remaining cells (median difference 0.005, p = 0.041).

  • BlockedFigure 4b and 4c· Data missing from the dataset

    During burst periods the overlap ensemble participates independently in more bursts than the aversive and neutral ensembles, with no difference between shock groups.

    During burst periods, the overlap ensemble participated independently in more bursts than the aversive ensemble ... and more than the neutral ensemble ... but there was no difference in participation across low- versus high-shock mice

    Figure 4b and 4c

    What the paper reports

    • overlap against aversive: t14 = 7.95, p = 2 × 10⁻⁶ (n = 7 low-shock and 8 high-shock mice)
    • overlap against neutral: t14 = 5.59, p = 10⁻⁴
    • low- against high-shock: F1,13 = 1.43, p = 0.25

    One-hour offline cohort not deposited.

  • BlockedFigure 4e and 4f; Extended Data Figure 4k to 4o· Data missing from the dataset

    In the one-hour offline experiment, high-shock mice show more overlap-neutral than overlap-aversive co-bursting during bursts (t7 = 4.95), low-shock mice do not, and there is no such difference outside bursts.

    Overlap ensemble preferentially co-participated with the neutral ensemble (N) rather than with the aversive ensemble (A) ... whereas in the low-shock group, there was no difference in overlap ensemble participation with the neutral and aversive ensembles

    Figure 4e and 4f; Extended Data Figure 4k to 4o

    What the paper reports

    • burst periods, ensemble combination by shock interaction (Figure 4e): F1,13 = 12.2, p = 0.004 (n = 7 low-shock and 8 high-shock mice)
    • burst periods, overlap-neutral against overlap-aversive, high-shock mice (Figure 4e): t7 = 4.95, p = 0.003
    • burst periods, overlap-neutral against overlap-aversive, low-shock mice (Figure 4e): t6 = 0.99, p = 0.36
    • non-burst periods, ensemble combination (Figure 4f): F1,13 = 0.027, p = 0.87
    • non-burst periods, interaction (Figure 4f): F1,13 = 1.11, p = 0.31

    One-hour offline cohort not deposited. The same analysis on the EEG cohort is tested in fig5c-wake-overlap-neutral-coburst.

  • BlockedExtended Data Figure 5; Extended Data Figure 6b and 6c· Data missing from the dataset

    The overlap ensemble is enriched in putative inhibitory neurons identified by chemotagging.

    Here, the Overlap ensemble comprised a larger fraction of the inhibitory neuron population than any of the other ensembles ... the Overlap ensemble was composed more of inhibitory neurons than the other ensembles were

    Extended Data Figure 5; Extended Data Figure 6b and 6c

    What the paper reports

    • fraction of the inhibitory population per ensemble (Extended Data Figure 6b): F3,12 = 26.17, p = 1.5 × 10⁻⁵ (n = 5 mice)
    • fraction of each ensemble that is inhibitory (Extended Data Figure 6c): F3,12 = 15.29, p = 2 × 10⁻⁴ (n = 5 mice)

    The inhibitory tagging (Gad2-cre, CNO) experiment is not in the dandiset.

  • BlockedExtended Data Figure 8d, 8f and 8g; Results, Co-bursting of the overlap and neutral ensemble, paragraph 4· Data missing from the dataset

    Without a shock, overlap-neutral co-bursting does not exceed overlap-aversive co-bursting and no memory linking occurs.

    the co-bursting of the overlap and neutral ensembles was not different from the co-bursting of the overlap and aversive ensembles during the offline period ... these mice showed no differences in freezing in the neutral versus novel contexts during recall

    Extended Data Figure 8d, 8f and 8g; Results, Co-bursting of the overlap and neutral ensemble, paragraph 4

    What the paper reports

    • co-bursting between ensemble pairs, normalized by chance (Extended Data Figure 8g): F2,10 = 1.29, p = 0.32 (n = 6 mice)
    • freezing, neutral against novel recall (Extended Data Figure 8d): t5 = 1.52, p = 0.19 (n = 6 mice)

    The no-shock cohort is not in the dandiset.

  • BlockedExtended Data Figure 7; Results, Co-bursting of the overlap and neutral ensemble, paragraph 4· Data missing from the dataset

    Neural activity during neutral and aversive encoding is highly discriminable, with no difference between low- and high-shock mice.

    We found that the neural activity patterns for neutral and aversive contexts were highly discriminable. There were no differences between the low- and high-shock groups during encoding

    Extended Data Figure 7; Results, Co-bursting of the overlap and neutral ensemble, paragraph 4

    What the paper reports

    • SVM accuracy, true against shuffled (Extended Data Figure 7a): F1,12 = 38.49, p = 4.6 × 10⁻⁵ (n = 6 low-shock and 8 high-shock mice)
    • SVM accuracy, low- against high-shock (Extended Data Figure 7a): F1,12 = 0.015, p = 0.22

    The one-hour offline cohort of Extended Data Figure 7 is not in the dandiset. Ca_EEG2-1's neutral and aversive encoding are registered, but the paper does not report this analysis for the EEG cohort.

  • BlockedExtended Data Figure 9c· Data missing from the dataset

    In the calcium imaging and EEG/EMG experiment, high-shock mice froze more in the neutral than in the novel context (t4 = 4.02, p = 0.03), while low-shock mice did not.

    High-shock mice froze more during Neutral vs Novel recall ... whereas low-shock mice froze no differently in Neutral vs Novel recall

    Extended Data Figure 9c

    What the paper reports

    • high-shock mice: t4 = 4.02, p = 0.03 (n = 5 high-shock mice)
    • low-shock mice: t3 = 1.08, p = 0.36 (n = 4 low-shock mice)

    The neutral and novel recall sessions of the only deposited high-shock mouse (Ca_EEG2-1) are not in the dandiset, and the paper's paired test needs several mice per group. The low-shock mouse's values (neutral 17.40%, novel 15.06%) are shown under source-data-link.

  • BlockedFigure 5c (middle and right)· Data missing from the dataset

    During NREM and REM sleep, overlap-neutral and overlap-aversive co-bursting do not differ in either shock group.

    Middle, for NREM, there was no difference in high-shock ... or low-shock ... co-bursting. Right, for REM, there was no difference in high-shock ... or low-shock ... co-bursting

    Figure 5c (middle and right)

    What the paper reports

    • NREM, high-shock mice: t4 = 0.53, p = 0.66 (n = 5 high-shock mice)
    • NREM, low-shock mice: t3 = -0.49, p = 0.66 (n = 4 low-shock mice)
    • REM, high-shock mice: t4 = 1.04, p = 0.63 (n = 5 high-shock mice)
    • REM, low-shock mice: t3 = -0.53, p = 0.63 (n = 4 low-shock mice)

    The three deposited offline recordings are each the first of 24 and hold only wake and quiet wake (no sws or rem interval); the later recordings, where sleep occurs, are not deposited.

  • BlockedFigure 5c (left)· Data missing from the dataset

    In low-shock mice, overlap-neutral and overlap-aversive co-bursting during wake do not differ (t3 = 1.20, P = 0.32).

    low-shock mice had no difference in co-bursting between these ensembles

    Figure 5c (left)

    What the paper reports

    • t3 = 1.20, p = 0.32 (n = 4 low-shock mice)

    The only deposited low-shock mouse (Ca_EEG3-4) has no cell registration in the dandiset, so its ensembles cannot be formed. With one mouse the paper's paired t-test could not be run either.

  • BlockedExtended Data Figure 9d to 9g and Extended Data Figure 10; Results, Co-reactivation occurs more during wake, paragraph 1· Data missing from the dataset

    Mice wearing a Miniscope sleep normally, the intermittent recording scheme captures the time spent in each state, and sleep structure is not changed by neutral or aversive encoding.

    We first validated that mice could wear the Miniscope without disruptions to their sleep ... We confirmed that our calcium recording scheme reliably captured all sleep states ... There were no differences in sleep features ... after the neutral or aversive encoding compared to pre-experiment sleep

    Extended Data Figure 9d to 9g and Extended Data Figure 10; Results, Co-reactivation occurs more during wake, paragraph 1

    What the paper reports

    • sleep duration, Miniscope against no Miniscope (Extended Data Figure 9d): F1,10 = 0.54, p = 0.48 (n = 8 Miniscope and 4 no-Miniscope mice)

    Needs continuous EEG/EMG and sleep scores for nine mice and a control group. Only Ca_EEG2-1 has week telemetry, without sleep scores; the other mice and the scores are not deposited.

  • BlockedFigure 6a; Extended Data Figure 4q to 4s; Results, Ensemble co-reactivation in neutral context recall, paragraph 2· Data missing from the dataset

    The overlap ensemble is reactivated more during neutral recall than during novel-context exposure in high-shock mice compared with low-shock mice (t12 = 3.2, P = 0.007).

    the cells that were active during both neutral and aversive encoding (overlap ensemble) were reactivated more during neutral recall than novel-context exposure in the high-shock but not the low-shock group

    Figure 6a; Extended Data Figure 4q to 4s; Results, Ensemble co-reactivation in neutral context recall, paragraph 2

    What the paper reports

    • overlap reactivation index, high- against low-shock (Figure 6a): t12 = 3.2, p = 0.007
    • overlap reactivation, neutral against novel, high-shock mice (Extended Data Figure 4s): t7 = 8.53, p = 1.2 × 10⁻⁴ (n = 8 high-shock mice)
    • overlap reactivation, neutral against novel, low-shock mice (Extended Data Figure 4s): t5 = 0.55, p = 0.61 (n = 6 low-shock mice)

    One-hour offline cohort not deposited; in the EEG cohort Ca_EEG2-1 lacks neutral and novel recall and Ca_EEG3-4 lacks registration.

  • BlockedFigure 6b· Data missing from the dataset

    In high-shock mice the neutral encoding-to-recall population correlation is lower than in low-shock mice and lower than the aversive encoding-to-recall correlation.

    During neutral recall in high-shock mice, population activity vectors were less correlated with the average neutral encoding population vector than aversive recall activity was with the average aversive encoding population vector ... Neutral encoding-to-recall correlations were also lower in high- versus low-shock mice

    Figure 6b

    What the paper reports

    • neutral against aversive encoding-to-recall correlation, high-shock mice: t7 = 4.10, p = 0.009 (n = 8 high-shock mice)
    • neutral encoding-to-recall correlation, high- against low-shock: t6,92 = 2.98, p = 0.042 (n = 6 low-shock and 8 high-shock mice)

    As fig6a-overlap-reactivated-neutral-recall; the neutral recall session needed is missing for the registered mouse.

Findings about the data

What attempts found about the dataset itself, for its owner and the archive.

  • Data missingThe dandiset holds two of the nine mice of the calcium imaging and EEG/EMG experiment (Ca_EEG2-1, high shock; Ca_EEG3-4, low shock) and, of each offline period, only the first 10 min recording of the 24 the Methods describe. Ca_EEG2-1 has no offline-day-1 recording and no neutral or novel recall session. The mice of the one-hour imaging and offline experiments (Figures 1d to 1g, 2 to 4 and 6, Extended Data Figures 3, 4 and 7), the behaviour-only experiments, and the inhibitory tagging and no-shock experiments are not deposited. Twenty of the 26 claims cannot be tested for this reason.
  • Data missingCa_EEG3-4 has no cell registration in the dandiset, so its ensembles cannot be formed. The Week file of Ca_EEG2-1 holds 19 registration tables for offline-day-2 recordings, of which only the first recording is deposited, and none for offline day 1.
  • Selection not documentedThe dandiset does not say why these two mice, and only the first offline recording of each period, were deposited, nor that the paper analysed more.
  • Metadata errorThe session start times of the Ca_EEG2-1 behaviour sessions (NeutralExposure, FC, Recall1), and the Week epochs built from them, are about 1 h later than the telemetry and Miniscope clock. As deposited, OfflineDay2Session1 starts 47 min before the FC session it follows. The telemetry is absent (mouse out of its home cage) in windows that contain each of these epochs moved 1 h earlier, and the offline file's EEG sits exactly at its own epoch in the week recording.
  • File format errorThe OnePhotonSeries timestamps of three files (Ca_EEG2-1 NeutralExposure, Ca_EEG3-4 OfflineDay1Session1, Ca_EEG3-4 Recall2) run at 16.7 ms and end at half the recording's length, while their frame counts fit 30 Hz over the full length given by the deconvolved timestamps.
  • Description missingThe sleep states include `quiet wake`, which the paper neither uses nor maps to its wake, NREM and REM states, so it is not known which state the paper assigned these epochs to. The Methods' scoring, applied to the EMG of Ca_EEG2-1, puts 94% of them in wake. The three offline recordings hold no NREM (`sws`) or REM epochs.
  • OtherThe telemetry (EEG, EMG, activity) is sampled at 500 Hz in the offline files and the Week file; the Methods say the HD-X02 implants recorded at 100 Hz.
  • Metadata errorThe session description of the offline recordings says that mouse behaviour and calcium were recorded for an hour with a webcam above the home cage. The deposited offline recordings last 10 min, hold telemetry and no video, and the Methods describe this experiment's offline recordings as 10 min in the telemetry home cage.
  • Description missingThe fluorescence series carry placeholder descriptions ("description of deconvolved traces") and unit "n.a.", the FC session descriptions give the shock amplitude without a unit, and the Week file has an empty session description.
  • Methods not availableThe analysis code repository the paper names (github.com/denisecailab/RetrospectiveMemoryLinkingAnalysis_2024) returns 404, so how bursts are counted, in the data and in shuffled surrogates, could not be checked; the surrogate burst rates here differ from the paper's chance level.

Notebooks

Every registered entry that uses this dataset. An entry can hold several notebooks, and can use more than one dataset.

Reproduction Offline ensemble co-reactivation in the EEG/EMG experiment (Zaki et al. 2024)

Tests the findings of Zaki et al. (2024) that can be tested on dandiset 000718, which holds two of the nine mice of the paper's calcium imaging and EEG/EMG experiment (one high-shock, one low-shock) with their encoding and recall sessions, the first offline recording after encoding, and cell registration for the high-shock mouse. It checks the deposited freezing against the paper's per-mouse values, compares freezing between the two mice, tests offline population bursts and their skew against shuffled activity and against movement, and compares overlap-neutral with overlap-aversive co-bursting in wake. The one-hour offline, behaviour-only, inhibitory tagging and no-shock experiments are not in the dandiset. All files

Verification

  • Round 1
    2026-10-18    Agreed with 25 of 26 outcomes. Independent agent run by CatalystNeuro (same harness, separate context, did not write the notebook). Reviewed: The executed notebook of the first run with the conclusion cells and the summary removed, the helper module and the claims file. The attempt record was not among its materials, so it took its counts from the claims file and the notebook.
    • offline-bursts-above-shuffle: the verifier’s outcome was “Not reproduced”. Not reproduced as the criterion is written, as the true burst rate is below the shuffled mean in all 3 recordings at z = 2. It did not count this as evidence against the paper, because z-scoring each surrogate again makes the test unable to detect coordination (in its synthetic data, 5.9 per min true against a null mean of 31.7 with planted coordination); the fixed-threshold variant behaves correctly but neither variant matches the paper's chance level. The criterion needs revising, or the claim should be treated as blocked with methods_underspecified. It asked for a positive control in the notebook.
  • Round 2
    2026-10-18    Agreed with 26 of 26 outcomes. Independent agent run by CatalystNeuro (same harness, separate context, did not write the notebook). Reviewed: Round 2. The full executed notebook of the revised run, with its conclusions and summary, the helper module, the claims file, the README and the attempt record. It read the paper from Europe PMC (PMC11666460) and did not re-run the notebook. Other points it raised, not tied to one claim, were corrected in the text; cell 10 now gives the random-pair centroid distances as 169 to 196 px.

Method differences

How the tests differed from the paper’s method or data, and the claims each applies to.

  • Only four per-mouse values can be compared, as the dandiset lacks the neutral and novel recall sessions of Ca_EEG2-1 and the other seven mice.source-data-link
  • One mouse per group (the paper has 5 high-shock and 4 low-shock mice); the test is within sessions, over 30 s bins, chosen here.ed9b-aversive-freezing-high-vs-low
  • Run on the three 10 min offline recordings of the EEG/EMG experiment at 15 Hz; the paper shows the test for the one-hour recordings at 30 Hz of the single-experience cohort (Extended Data Figure 3), which is not deposited.offline-bursts-above-shuffle
  • Burst events taken as maximal runs of frames above threshold, as the Methods do not define an event further.offline-bursts-above-shuffle
  • Run on the EEG/EMG experiment's three 10 min offline recordings; the paper shows the test for another cohort (Extended Data Figure 3o), which is not deposited.offline-skew-above-shuffle
  • Movement is the magnitude of the HD-X02 telemetry activity signal; the paper tracked locomotion with a webcam, whose video is not deposited.burst-locomotion-pause
  • Tested on the EEG/EMG experiment's offline recordings; the paper shows the effect for the single-experience cohort (Figure 2d and 2e) and the inhibitory tagging cohort (Extended Data Figure 6e).burst-locomotion-pause
  • One high-shock mouse of five, and its first offline recording of 24 (10 min of 4 h).fig5c-wake-overlap-neutral-coburst
  • The Methods z-score the ensemble traces over all 24 offline recordings concatenated; here only one is deposited, so they are z-scored within it.fig5c-wake-overlap-neutral-coburst
  • The criterion fixed four readings (quiet wake in or out of wake; co-participation with or without the third ensemble) and would count the claim under any of them. After the first verification the outcome is judged on wake including quiet wake, as the Methods' own EMG scoring puts quiet wake in wake; under wake alone OxN exceeds OxA without a significant difference, which the criterion would have called partially reproduced.fig5c-wake-overlap-neutral-coburst
  • The EMG feature of the Methods' scoring (log RMS per 6 s epoch) is chosen here, after the data were seen; the Methods do not name one. Without the log the mixture does not separate sleep from wake.fig5c-wake-overlap-neutral-coburst
  • Bursts are assigned the sleep state of their peak frame, a choice made here.fig5c-wake-overlap-neutral-coburst

Provenance

Attempt record
000718-zaki-2024-01
Paper
Offline ensemble co-reactivation links memories across days
Date
2026-10-18
Made by
An agent, run by CatalystNeuro
Dataset
DANDI:000718
Version
0.260825.1902 (doi:10.48324/dandi.000718/0.260825.1902)
Run on
GitHub Actions
Harness
Claude Code (GitHub Actions)
Model
claude-opus-5-5
Protocol
pilot-7

Transcripts of the agent runs

Each run’s prompt, messages and tool calls, with long output cut and the text of the paper left out, and its final report.

← DANDI:000691DANDI:000727 →