Datasets / DANDI:000061

DANDI:000061

✓ Verified by an independent agent

Reactivations of emotional memory in the hippocampus–amygdala system during sleep

electrophysiological Brown rat Converted by CatalystNeuro 2.0 TB 40 files
  • 6 reproduced
  • 1 partially reproduced
  • 3 not reproduced
  • 22 blocked

View on DANDI Paper JSON How to read outcomes

Claims

From the paper Reactivations of emotional memory in the hippocampus–amygdala system during sleep

Reproducing Girardeau et al. (2017) from dandiset 000061 10

Notebook reproduce_girardeau_2017.ipynb

  • Not reproducedResults, BLA recordings and sleep physiology, paragraph 1; Supplementary Table 1

    Over 61 sessions in 4 rats, 7,390 well-isolated units were recorded (rat 1, 2,444; rat 2, 1,294; rat 3, 1,138; rat 4, 2,514), of which 2,038 were in the BLA, 782 in the central nuclei, 1,560 in the piriform cortex and 1,210 in the hippocampus.

    Over the course of a total of 61 sessions, we recorded 7,390 well-isolated units (rat 1, 2,444; rat 2, 1,294; rat ... 3, 1,138; rat 4, 2,514). On the basis of histological reconstruction of probe placement and probe movement record, 2,038 of these were in the BLA, 782 in the central nuclei, 1,560 in the piriform cortex and 1,210 in the hippocampus

    Results, BLA recordings and sleep physiology, paragraph 1; Supplementary Table 1

    The dandiset holds 40 sessions (33 with a units table) and 4702 units, against 61 sessions and 7,390 units, both outside the 10% tolerance, so the criterion's totals fail. Only Rat08, taken as rat 1, is within tolerance (2317 against 2,444); Rat09, Rat10 and Rat11 hold 515, 440 and 1430 against 1,294, 1,138 and 2,514. By label there are 1118 BLA (bla and BLA), 482 central nuclei, 935 piriform and 829 hippocampal units against 2,038, 782, 1,560 and 1,210; the wider amygdala labels (bla, blv, bmp, ladl) give 1977. The difference is in the data deposited, fewer days and units than the paper reports. No outcome in the earlier attempt, which had no such claim.

    How the test differed from the paper’s method

    • The dandiset does not say which of Rat08 to Rat11 are the paper's rats 1 to 4. Rat09, the rat without hippocampal units, is taken as rat 2 (whose hippocampal probe failed) and the others in order; this is an assumption.
    • Units are counted as stored, one per session, including units 0 and 1 of the ten files whose spike index is wrong; location counts use the dandiset's labels read as the paper's structures.
  • Not reproducedFigure 2c legend

    BLA interneurons do not change their firing rate between REM sleep and wake, and fire slightly less in NREM sleep than in wake.

    Interneurons do not change firing rates between REM and wake ... and slightly decrease firing rates during NREM compared to wake

    Figure 2c legend

    What the paper reports

    • REM/wake firing rate ratio of BLA interneurons: Wilcoxon signed rank test, z = 1.25, p = 0.21 (n = 175 monosynaptically identified interneurons)
    • NREM/wake firing rate ratio of BLA interneurons: Wilcoxon signed rank test, z = -6.00, p = 2 × 10⁻⁹ (n = 175 monosynaptically identified interneurons)

    On the 36 identified inhibitory BLA units of five Rat08 sessions, the REM/wake ratio has a median of 0.029 (positive in 22) with two-sided Wilcoxon p = 0.020, so the paper's test finds a REM change the claim says is absent. The NREM/wake ratio has a negative median (-0.035, positive in 16) with p = 0.26, so the decrease is not significant. Neither part holds, which the criterion calls not reproduced. By session only Rat08-20130711 (12 cells) has a REM ratio with p < 0.05. The paper had 175 interneurons from four rats; the REM effect here is small (median 0.029) and rests on one rat. The answer to this attempt's question (questions/000061.yaml) confirmed that the claim is judged on O4 reading a alone; the outcome, not reproduced before the answer, is unchanged. As context (decides nothing), the REM effect size r is 0.39 against the paper's 0.09 and the bootstrap 95% interval of the REM median is -0.015 to 0.109; the NREM r is -0.19 against the paper's -0.45, and with 36 cells a NREM shift of about 0.068 would be detected with 80% power. A short-latency cross-correlogram check finds an inhibitory partner at p < 1e-6 for 34 of the 36 inhibitory units (94.4%), against 2.0% of excitatory and 7.9% of unclassified BLA units, consistent with the column marking identified interneurons. No outcome in the earlier attempt.

    How the test differed from the paper’s method

    • Cells are the units marked inhibitory in the five typed Rat08 files, read as the paper's identified interneurons (O4 reading a, which as answered decides alone); reading b cannot be formed.
  • Not reproducedResults, A subset of BLA cells are modulated during hippocampal SPW-Rs, paragraph 1; Supplementary Fig. 5b

    Reactivation (EV - REV) is larger for hippocampus-BLA pairs whose BLA cell is ripple-modulated than for pairs with a non-modulated BLA cell.

    Moreover, reactivations calculated using SPW-R-modulated BLA cells were larger than for nonmodulated pairs (Supplementary Fig. 5b).

    Results, A subset of BLA cells are modulated during hippocampal SPW-Rs, paragraph 1; Supplementary Fig. 5b

    What the paper reports

    • EV - REV, ripple-modulated against non-modulated BLA partners, all cells: Wilcoxon ranksum test, z = 2.71, p = 0.00662 (n = 19 sessions in 3 rats)
    • EV - REV, ripple-modulated against non-modulated BLA partners, pyramidal pairs: Wilcoxon ranksum test, z = 2.57, p = 0.00992 (n = 18 sessions in 3 rats)

    With all cells over 12 sessions (the paper had 19), EV - REV is larger with the modulated BLA units in 7, mean 4.18% against 1.68% for the non-modulated units, two-sided rank sum p = 0.27 under O2 reading a and 0.3 under reading b, so the all-cell test does not hold. The non-modulated mean is within the expected 1.76 +/- 0.56%; the modulated mean is below the expected 6.07 +/- 1.3%. By rat the modulated units give the larger value in 4 of 7 Rat08, 0 of 2 Rat10 and 3 of 3 Rat11 sessions; 12 of the 23 sessions with sleep and ripples have more than 5 hippocampal and 5 modulated BLA units. The pyramidal test cannot be run (no putative types). Before the answer to this attempt's question (questions/000061.yaml) the claim was blocked as metadata_insufficient through bla-ripple-up-subgroup; the answer removed that dependency, M6 needing no cell type, and the criterion now judges the all-cell test alone when the pyramidal test cannot be run, not reproduced when it does not hold. No outcome in the earlier attempt.

    How the test differed from the paper’s method

    • Only the all-cell test is run; the pyramidal-pair test needs putative types the dandiset lacks.
    • BLA is the bla label (case ignored); units 0 and 1 of the files with a wrong spike index are left out.
  • Partially reproducedSupplementary Fig. 11a

    Reactivation strength of the whole-run template around hippocampal ripple peaks is higher in post-NREM than in pre-NREM, for pyramidal pairs and for all cell pairs.

    Mean z-scored peri-ripple reactivation strength of the whole training session ... reactivation strength in a 500ms window around ripple peaks (grey bar in a) was significantly higher in post-NREM

    Supplementary Fig. 11a

    What the paper reports

    • post-NREM against pre-NREM ripple-peak reactivation strength, whole training session, pyramidal cell pairs: Wilcoxon one-tailed signed rank test, z = -3.175, p = 7.5 × 10⁻⁴ (n = 25 sessions in 3 rats)
    • the same, all cell pairs: Wilcoxon one-tailed signed rank test, z = -2.895, p = 0.00189 (n = 27 sessions in 3 rats)

    With all cells, the ripple-peak strength of the whole-run template is higher in post-NREM than in pre-NREM in 11 of 15 sessions (median difference 0.0224, one-tailed p = 0.0042; the paper 0.00189 over 27 sessions); post exceeds pre in 7 of 9 Rat08, 2 of 3 Rat10 and 2 of 3 Rat11 sessions. The pyramidal test cannot be run, so the pyramidal part of the statement is not tested. R leaves out units silent in the template or in a sleep epoch's NREM, 5 of all units entering (1 hippocampal in Rat08-20130713, 2 in Rat11-20150326, 2 BLA in Rat11-20150325). Before the answer to this attempt's question (questions/000061.yaml) the claim was blocked as metadata_insufficient through hpc-bla-reactivation-post-sleep; the answer removed that dependency, sessions entering by the M7 cell minimum alone, and the criterion now judges the all-cell test alone when the pyramidal test cannot be run, partially reproduced when it has p < 0.05. The earlier attempt reproduced this id with singular-vector patterns and two readings of the abstract; the claim now uses the paper's reactivation strength R.

    How the test differed from the paper’s method

    • Only the all-cell test is run; the pyramidal test needs putative types the dandiset lacks.
    • BLA is the bla label (case ignored); units 0 and 1 of the files with a wrong spike index are left out.
  • ReproducedResults, Rats learn the daily location of an aversive air puff on a linear track, paragraph 2; Supplementary Fig. 1c

    The aversive valence of the air puff diminishes over training days; rats run more laps per unit time as they habituate.

    The aversive valence of the air puff gradually diminished with training days (Supplementary Fig. 1c).

    Results, Rats learn the daily location of an aversive air puff on a linear track, paragraph 2; Supplementary Fig. 1c

    Under O1 reading c, over 40 sessions of four rats, the habituation index rises with training day for each lap count, pooled Spearman rho 0.383 (p = 0.015) for all laps, 0.331 (p = 0.046, 37 sessions with a puff direction) for air puff laps and 0.397 (p = 0.011) for left-to-right laps. The correlation is positive in all four rats for each count, with p < 0.05 in Rat08, Rat09 and Rat10 and not in Rat11 (rho 0.32 to 0.33). Both readings of "gradually diminished" hold for all three counts, so the outcome holds under every reading the dandiset allows. The answer to this attempt's question (questions/000061.yaml) confirmed that the claim is judged on O1 reading c alone when the dataset holds neither lap intervals nor position, and the criterion now says so; the outcome, reproduced before the answer, is unchanged. No outcome in the earlier attempt.

    How the test differed from the paper’s method

    • Laps are O1 reading c, the intervals between consecutive rewards at opposite ends with both rewards inside the run. Reading a has no lap intervals in the dandiset (its trials table is not laps) and reading b needs position, which it lacks. As answered, the claim is judged on reading c alone; the hand-picked 150-cm window of ClassifyLaps.m is not reproduced by any reading. A reading c lap runs from reward to reward, so it includes the time on the starting platform and any stop, which the M8 laps leave out; the count of laps is the same either way.
    • The air puff direction of a day is the lap direction holding most puffs (1153 of 1197 puffs fall in run laps of that direction); the dandiset gives no direction. Rat08-20130709 has no puff inside a run lap and has no air puff count.
    • Training day is the rank of the recording date within each rat, over the 40 files; the paper's 55 behaviour sessions include days the dandiset does not hold.
  • ReproducedResults, BLA recordings and sleep physiology, paragraph 1; Figure 2c

    BLA pyramidal cells fire faster in REM sleep and, to a lesser extent, in NREM sleep than in wakefulness.

    we found a specific increase in the firing rate of BLA pyramidal cells, but not interneurons, during REM sleep and, to a lesser extent, NREM sleep relative to wakefulness

    Results, BLA recordings and sleep physiology, paragraph 1; Figure 2c

    What the paper reports

    • NREM/wake firing rate ratio of BLA pyramidal cells: Wilcoxon signed rank test, z = 14.25, p = 4.1 × 10⁻⁴⁶ (n = 675 monosynaptically identified pyramidal cells)
    • REM/wake firing rate ratio of BLA pyramidal cells: Wilcoxon signed rank test, z = 13.54, p = 8 × 10⁻⁴² (n = 675 monosynaptically identified pyramidal cells)

    On the 152 identified excitatory BLA units of five Rat08 sessions, the REM/wake ratio has a median of 0.39 (positive in 120, p = 7.0e-15) and the NREM/wake ratio a median of 0.18 (positive in 108, p = 7.2e-12); both are positive and the REM median exceeds the NREM median. By session the REM ratio reaches p < 0.05 in four of five sessions and the NREM ratio in three (Rat08-20130713, 22 cells, reaches neither). The paper had 675 cells from four rats; these 152 come from one rat, a smaller sample. The answer to this attempt's question (questions/000061.yaml) confirmed that the claim is judged on O4 reading a alone, the identified cells of the Fig. 2c legend, and O4 now says so; the outcome, reproduced before the answer, is unchanged. No outcome in the earlier attempt.

    How the test differed from the paper’s method

    • Cells are the units the dandiset's cell_type column marks excitatory, read as the paper's monosynaptically identified pyramidal cells (O4 reading a). The column exists in five Rat08 files only, so the test uses one rat and about a fifth of the paper's cells. Reading b (all putative pyramidal cells) cannot be formed; as answered, it is not needed for the outcome.
    • Units 0 and 1 of files with a wrong spike index are left out (Rat08-20130710 and 20130711).
  • ReproducedResults, BLA-hippocampus coordinated reactivations during NREM sleep, paragraph 1; Supplementary Fig. 5a

    Hippocampus-BLA NREM reactivation holds when pyramidal cells and interneurons are both included.

    Reactivations were maintained when both pyramidal cells and interneurons were included in the EV calculation (Supplementary Fig. 5).

    Results, BLA-hippocampus coordinated reactivations during NREM sleep, paragraph 1; Supplementary Fig. 5a

    What the paper reports

    • EV against REV, NREM, all cell pairs: Wilcoxon signed rank test, z = 3.79, p = 1.4 × 10⁻⁴ (n = 27 sessions in 3 rats)
    • EV against REV, 0-20 min of post-NREM, all cell pairs: Wilcoxon signed rank test, z = 3.91, p = 8.8 × 10⁻⁵ (n = 17 sessions in 3 rats)
    • EV against REV, 20-40 min, all cell pairs: Wilcoxon signed rank test, z = 2.98, p = 0.00282 (n = 17 sessions in 3 rats)
    • EV against REV, 40-60 min, all cell pairs: Wilcoxon signed rank test, z = -0.71, p = 0.14458 (n = 17 sessions in 3 rats)

    Over 15 sessions (9 Rat08, 3 Rat10, 3 Rat11; the paper had 27), EV exceeds REV in 13, mean EV - REV 3.48% (s.e.m. 0.99), two-sided Wilcoxon p = 0.00031. EV exceeds REV in all 9 Rat08 and all 3 Rat11 sessions and in 1 of 3 Rat10 sessions (median -0.023%). The wider amygdala labels give 22 sessions, 17 with EV above REV, p = 0.0005. Reported only, EV - REV in successive 20-min periods of post-NREM has means of 2.49, 1.27 and 0.52% (p = 0.018, 0.035 and 1). No outcome in the earlier attempt.

    How the test differed from the paper’s method

    • BLA is the bla label (case ignored); the wider amygdala labels are a variant.
    • Units 0 and 1 of the files with a wrong spike index are left out.
    • A 50-ms bin belongs to an epoch when its centre does, a standard choice.
  • ReproducedResults, BLA-hippocampus coordinated reactivations during NREM sleep, paragraph 2; Figure 3e

    There are no reactivations between the central amygdala nuclei and the hippocampus.

    while there were no reactivations at all between the central nuclei and the hippocampus (Fig. 3e)

    Results, BLA-hippocampus coordinated reactivations during NREM sleep, paragraph 2; Figure 3e

    What the paper reports

    • EV against REV, NREM, all cell pairs: Wilcoxon signed rank test, z = 0.36, p = 0.717 (n = 19 sessions in 3 rats)
    • EV against REV, REM, all cell pairs: Wilcoxon signed rank test, z = -0.40, p = 0.687 (n = 19 sessions in 3 rats)

    Hippocampus-central nuclei EV does not differ from REV in NREM (9 sessions, median EV - REV 0.60%, EV above REV in 7, p = 0.30) or in REM (8 sessions, median 0.085%, 6 of 8, p = 0.25), so the paper's test finds no reactivation. The paper had 19 sessions in each; with 9 and 8 the test has less power, and the NREM median is positive. As context (decides nothing), the effect sizes r are 0.34 (NREM) and 0.41 (REM) against the paper's 0.08 and -0.09; with these sessions a shift of the median EV - REV of about 3.0% (NREM) or 0.95% (REM) would be detected with 80% power, and the bootstrap 95% interval of the NREM median, -2.4 to 4.1%, includes the hippocampus-BLA median of 1.96%. Resampling the hippocampus-BLA sessions at 9 and 8, a reactivation distributed as the BLA's reaches p < 0.05 in 92% and 93% of samples. The absence holds on the paper's test and is weak evidence. No outcome in the earlier attempt.

    How the test differed from the paper’s method

    • Central nuclei are the cen label. Units 0 and 1 of files with a wrong spike index are left out.
  • ReproducedSupplementary Fig. 11b,c

    With all cell types included, both the air puff and the safe trajectory patterns are more reactivated around post-NREM than pre-NREM ripples, and the post minus pre difference does not differ significantly between the two trajectories.

    Reactivation strength in a 500ms window around ripple peaks (grey bar in a) was significantly higher in post-NREM compared to pre-NREM for airpuff ... and safe trajectories ... Pre/Post peak differences (postNREM - preNREM; see methods) in reactivation strength are not significantly different for safe and airpuff trajectories when all cell types are included

    Supplementary Fig. 11b,c

    What the paper reports

    • post-NREM against pre-NREM, air puff trajectory, all cell pairs: Wilcoxon one-tail signed rank test, z = -3.327, p = 4.4 × 10⁻⁴ (n = 27 sessions in 3 rats)
    • post-NREM against pre-NREM, safe trajectory, all cell pairs: Wilcoxon one-tail signed rank test, z = -1.789, p = 0.036 (n = 27 sessions in 3 rats)
    • post minus pre difference, air puff against safe, all cell pairs: Wilcoxon one-tailed signed rank test, z = -1.333, p = 0.0912 (n = 27 sessions in 3 rats)

    With all cells over 14 sessions (the paper had 27), the air puff template is more reactivated around post-NREM than pre-NREM ripples in 10 (one-tailed p = 0.01), the safe template in 12 (p = 0.0043), and the air puff change exceeds the safe change in 8 (one-tailed p = 0.48), so both within-trajectory tests have p < 0.05 and the between-trajectory test has p >= 0.05. By rat the air puff template rises in 7 of 8 Rat08, 2 of 3 Rat10 and 1 of 3 Rat11 sessions, so its result rests mostly on Rat08. Reading c laps include the time on the starting platform and stops (median lap 22.6 s, but minutes long in early Rat08 sessions); leaving out Rat08-20130708, the only session with fewer than 3 laps in a template (one safe lap of 23 s), gives air puff p = 0.0052, safe p = 0.0052 and air puff greater than safe p = 0.37 (13 sessions). Long laps are not limited to that session (median air puff laps of 660 s in Rat08-20130711, 192 s in Rat08-20130710 and 46 to 115 s in four other Rat08 sessions), so two further variants, with values of the author's that decide nothing, limit the platform time and stops. With laps of at most 30 s (11 sessions) the result is the same (air puff p = 0.034, safe p = 0.021, air puff greater than safe p = 0.71). With each lap cut to its last 10 s (14 sessions) the air puff test holds (p = 0.034) but the safe test does not reach p < 0.05 (10 of 14, p = 0.052), and air puff greater than safe has p = 0.36. The result under the criterion's laps thus depends in part on the time around the run that reading c laps hold; the criterion fixes reading c, so the outcome is unchanged. Before the answers to this attempt's questions (questions/000061.yaml) the claim was blocked as metadata_insufficient through aversive-direction-stronger; the answers removed that dependency, the claim using that claim's procedure but not its result, and confirmed that laps are judged on O1 reading c alone here. No outcome in the earlier attempt.

    How the test differed from the paper’s method

    • Laps are O1 reading c (between rewards at opposite ends); the dandiset has no lap intervals or position. The air puff direction of a day is the lap direction holding most puffs. These laps include the time on the starting platform after each reward and any stop, so the air puff and safe templates hold awake activity off the track that the M8 laps (position inside a 150-cm window) leave out.
    • BLA is the bla label (case ignored); units 0 and 1 of the files with a wrong spike index are left out.
  • ReproducedResults, The aversive trajectory is reactivated during SPW-Rs, paragraph 1; Supplementary Fig. 11d

    The occurrence rate of SPW-Rs does not differ between pre-NREM and post-NREM.

    The occurrence rate of SPW-R was also not significantly different between the pre-NREM and postNREM epochs

    Results, The aversive trajectory is reactivated during SPW-Rs, paragraph 1; Supplementary Fig. 11d

    What the paper reports

    • Wilcoxon signed rank test, z = -1.289, p = 0.192 (n = 41 sessions in 3 rats)

    Over 23 sessions (the paper had 41), the median ripple rate is 0.611 Hz in pre-NREM and 0.587 Hz in post-NREM; post minus pre has a median of 0.0315 Hz, is positive in 15, and the two-sided Wilcoxon p is 0.26, so the paper's test finds no difference. By rat the median change is 0.089 Hz in Rat08 (8 of 9 higher in post), -0.008 Hz in Rat10 and 0.007 Hz in Rat11. As context (decides nothing), the effect size r is 0.24 against the paper's 0.20, the bootstrap 95% interval of the median change is -0.008 to 0.051 Hz (-1.4% to 8.4% of the median pre-NREM rate of 0.611 Hz), and 23 sessions detect a change of about 0.068 Hz (11% of the rate) with 80% power (0.050 Hz with the paper's 41); a change larger than about a tenth of the rate is excluded, a smaller one is not. No outcome in the earlier attempt.

    How the test differed from the paper’s method

    • Ripples are the dandiset's rip annotations, counted in NREM when their peak is, a standard choice.

Not tested in a notebook 22

5 too large to analyse, 17 metadata insufficient
  • BlockedResults, Rats learn the daily location of an aversive air puff on a linear track, paragraph 1; Figure 1b, right; Supplementary Fig. 1b· Too large to analyse

    Rats learned the daily location of the air puff. The speed ratio between the previous and the current danger zone rises from pre-run to run and post-run, reversing between pre-run and post-run.

    Therefore, the systematic reversal in the speed ratio (previous/current danger zone) between pre- and ... post-run indicates learning of the new air puff location (Fig. 1b and Supplementary Fig. 1b).

    Results, Rats learn the daily location of an aversive air puff on a linear track, paragraph 1; Figure 1b, right; Supplementary Fig. 1b

    What the paper reports

    • speed ratio over pre-run, run and post-run: one-way repeated measure ANOVA, F = 40.48, d.f. = 2, p = 1.3 × 10⁻¹³ (n = 55 sessions in 4 animals)
    • speed ratio, pre-run against run: post hoc t-test, t(51) = -8.49, p = 2.5 × 10⁻¹¹, Bonferroni (on the significance thresholds; the P shown is uncorrected)-corrected
    • speed ratio, pre-run against post-run: post hoc t-test, t(51) = -5.08, p = 5.3 × 10⁻⁶, Bonferroni (on the significance thresholds; the P shown is uncorrected)-corrected
    • speed ratio, run against post-run: post hoc t-test, t(50) = 4.43, p = 5.1 × 10⁻⁵, Bonferroni (on the significance thresholds; the P shown is uncorrected)-corrected

    The criterion needs the rat's speed in the 20 cm before the day's and the previous day's air puff location, so it needs position on the track and the puff locations. The dandiset holds neither, only puff times and the raw video frames (40 run videos, 1920573 frames, 1328 GB uncompressed); streaming 20 frames took 8.3 s in the final run, about 222 hours for the run frames alone, before any tracking, and the video timestamps are not on the session clock. No file has a Position, SpatialSeries or other behavioural time series group (the data check lists the groups of every file). Same outcome as the earlier attempt, which blocked it for the same reason.

  • BlockedSupplementary Fig. 1b legend· Too large to analyse

    The speed ratio changes across pre-run, run and post-run in each of the four rats taken alone, indicating that every rat learned the new air puff location.

    The reversal of ratio between pre and post RUN indicates that all animals learned the new location during RUN and remembered it during post-RUN

    Supplementary Fig. 1b legend

    What the paper reports

    • speed ratio over pre-run, run and post-run, rat 1: one-way repeated measure ANOVA, F = 13.26, d.f. = 2, p = 2.1 × 10⁻⁴ (n = 11 sessions)
    • the same, rat 2: one-way repeated measure ANOVA, F = 7.02, d.f. = 2, p = 0.00556 (n = 10 sessions)
    • the same, rat 3: one-way repeated measure ANOVA, F = 27.41, d.f. = 2, p = 3.9 × 10⁻⁷ (n = 14 sessions)
    • the same, rat 4: one-way repeated measure ANOVA, F = 4.36, d.f. = 2, p = 0.021754 (n = 16 sessions)

    Depends on rats-learn-puff-location (speed ratios from position), which is blocked as too large. No outcome in the earlier attempt.

  • BlockedResults, Rats learn the daily location of an aversive air puff on a linear track, paragraph 2; Figure 1b, left; Figure 1c· Too large to analyse

    Speed in the current danger zone is lower in post-run than in pre-run, and in run and post-run rats run slower in the current than in the previous danger zone.

    This was quantitatively reflected by the significantly slower speed in the current danger zone in post-run compared to pre-run (Fig. 1b,c).

    Results, Rats learn the daily location of an aversive air puff on a linear track, paragraph 2; Figure 1b, left; Figure 1c

    What the paper reports

    • speed over epochs (pre-run, run, post-run): two-way repeated measures ANOVA, session effect, F = 41.51, d.f. = 2, p = 7.5 × 10⁻¹⁴ (n = 55 sessions in 4 animals)
    • speed, current against previous danger zone: two-way repeated measures ANOVA, air puff location effect, F = 11.81, d.f. = 1, p = 0.0012
    • epoch by danger zone: two-way repeated measures ANOVA, interaction, F = 30.53, d.f. = 2, p = 4.5 × 10⁻¹¹
    • previous against current danger zone speed, pre-run: post hoc paired t-test, t(52) = 3.45, p = 0.0011, Bonferroni-corrected
    • previous against current danger zone speed, run: post hoc paired t-test, t(51) = -6.84, p = 9.7 × 10⁻⁹, Bonferroni-corrected
    • previous against current danger zone speed, post-run: post hoc paired t-test, t(51) = -3.10, p = 0.00307, Bonferroni-corrected
    • current danger zone speed, pre-run against post-run: post hoc paired t-test, t(53) = 5.47, p = 1.2 × 10⁻⁶, Bonferroni-corrected

    Depends on rats-learn-puff-location (danger-zone speeds from position), which is blocked as too large. No outcome in the earlier attempt.

  • BlockedResults, Rats learn the daily location of an aversive air puff on a linear track, paragraph 2· Too large to analyse

    The location of the air puff on the track does not correlate with the speed in the current danger zone in pre-run, run or post-run.

    The location of the air puff on the track did not correlate with the speed in the current danger zone in pre-run, training or post-run

    Results, Rats learn the daily location of an aversive air puff on a linear track, paragraph 2

    Depends on rats-learn-puff-location and needs the puff location on the track, which the dandiset does not hold (only puff times). No outcome in the earlier attempt.

  • BlockedResults, BLA-hippocampus coordinated reactivations during NREM sleep, paragraph 1; Figure 3a-c· Metadata insufficient

    Hippocampus-BLA pyramidal cell pair correlations of the run are reinstated in post-sleep NREM, measured as explained variance (EV) exceeding the reverse explained variance (REV) across sessions.

    The EV and REV, calculated using hippocampus–BLA pyramidal cell pairs, showed significant reactivations between the hippocampus and BLA during post-NREM (Fig. 3c).

    Results, BLA-hippocampus coordinated reactivations during NREM sleep, paragraph 1; Figure 3a-c

    What the paper reports

    • EV against REV, NREM, pyramidal pairs: Wilcoxon signed rank test, z = 3.72, p = 1.9 × 10⁻⁴ (n = 25 sessions in 3 rats)

    The criterion computes EV and REV over hippocampus-BLA putative pyramidal pairs. The dandiset has no putative cell types; its cell_type column (five Rat08 files) marks only identified cells, and no file has six hippocampal units marked excitatory. The earlier attempt reproduced this id with another measure (singular-vector joint patterns over all units at 0.1 Hz or more); the claim now uses the paper's EV against REV on pyramidal pairs, which the dandiset cannot select. The same test with all cells is reactivation-all-cell-pairs, reproduced here (15 sessions, p = 0.00031).

  • BlockedResults, BLA-hippocampus coordinated reactivations during NREM sleep, paragraph 1; Figure 3c, right· Metadata insufficient

    Hippocampus-BLA reactivation decays over the first hour of post-sleep NREM (mean EV - REV 2.89, 1.72 and 1.31% in successive 20-min NREM periods).

    The gradual decay in reactivations over the first hour of NREM sleep ... paralleled the previously described decay for pairs of hippocampal neurons

    Results, BLA-hippocampus coordinated reactivations during NREM sleep, paragraph 1; Figure 3c, right

    What the paper reports

    • EV - REV across the three 20-min NREM periods: Kruskal-Wallis test, chi2 = 8.53, d.f. = 2, p = 0.014 (n = 19 sessions)
    • EV against REV, 0-20 min of post-NREM: Wilcoxon signed rank test, z = 0.38 (as printed), p = 1.3 × 10⁻⁴ (n = 19 sessions in 3 rats)
    • EV against REV, 20-40 min: Wilcoxon signed rank test, z = 3.5, p = 4.6 × 10⁻⁴ (n = 19 sessions in 3 rats)
    • EV against REV, 40-60 min: Wilcoxon signed rank test, z = 2.15, p = 0.0312 (n = 19 sessions in 3 rats)

    Pyramidal pairs, and depends on hpc-bla-reactivation-post-sleep, which is blocked. The all-cell 20-min periods are reported under reactivation-all-cell-pairs (means 2.49, 1.27 and 0.52%). No outcome in the earlier attempt.

  • BlockedResults, BLA-hippocampus coordinated reactivations during NREM sleep, paragraph 2; Figure 3c, left· Metadata insufficient

    Hippocampus-BLA and hippocampus-piriform cortex reactivations are not significant in REM sleep.

    Reactivations during REM sleep ... were not significant in any structure

    Results, BLA-hippocampus coordinated reactivations during NREM sleep, paragraph 2; Figure 3c, left

    What the paper reports

    • EV against REV, REM, pyramidal pairs: Wilcoxon signed rank test, z = 0.88, p = 0.377 (n = 23 sessions in 3 rats)
    • EV against REV, REM, all cell pairs (Supplementary Fig. 5a): Wilcoxon signed rank test, z = -0.71, p = 0.47583 (n = 25 sessions in 3 rats)
    • EV against REV, REM, hippocampus-piriform pyramidal pairs (Fig. 3d): Wilcoxon signed rank test, p = 0.301 (n = 15 sessions in 3 rats)

    Both tests of the criterion use pyramidal pairs (hippocampus-BLA and hippocampus-piriform). The criterion's variant, hippocampus-BLA with all cells in REM, is shown as a description, 14 sessions, EV above REV in 8, p = 0.43. No outcome in the earlier attempt.

  • BlockedResults, BLA-hippocampus coordinated reactivations during NREM sleep, paragraph 2; Figure 3d· Metadata insufficient

    Piriform cortex neurons show weaker reactivation with hippocampal CA1 neurons than BLA neurons do, though still significant in NREM.

    Neurons in the piriform cortex showed weaker experience-induced ... reactivation with their CA1 partner neurons (Fig. 3d

    Results, BLA-hippocampus coordinated reactivations during NREM sleep, paragraph 2; Figure 3d

    What the paper reports

    • EV - REV, hippocampus-BLA against hippocampus-piriform: Wilcoxon one-tailed rank sum test, z = 2.38, p = 0.0092 (n = 25 hippocampus-BLA and 14 hippocampus-piriform sessions)
    • EV against REV, hippocampus-piriform, NREM, pyramidal pairs: Wilcoxon signed rank test, p = 0.0052 (n = 17 sessions)

    Pyramidal pairs, and depends on hpc-bla-reactivation-post-sleep, which is blocked. No outcome in the earlier attempt.

  • BlockedSupplementary Fig. 4; Results, BLA-hippocampus coordinated reactivations during NREM sleep, paragraph 1· Metadata insufficient

    Within the hippocampus and within the BLA, pair correlations are reactivated in NREM but not in REM sleep.

    Not quoted: No sentence states this finding. It is read from the tests in the Supplementary Fig. 4 legend (titled "Intra-structure reactivations for hippocampus and BLA"), with the REM part also covered by the Results sentence that REM reactivations were not significant in any structure.

    What the paper reports

    • hippocampus, all pairs, NREM: Wilcoxon signed rank test, z = 3.95, p = 7.7 × 10⁻⁵ (n = 38 sessions in 3 rats)
    • hippocampus, pyramidal pairs, NREM: Wilcoxon signed rank test, z = 3.99, p = 6.4 × 10⁻⁵ (n = 38 sessions in 3 rats)
    • BLA, all pairs, NREM: Wilcoxon signed rank test, z = 2.74, p = 0.00606 (n = 34 sessions in 4 rats)
    • BLA, pyramidal pairs, NREM: Wilcoxon signed rank test, z = 3.16, p = 0.00157 (n = 32 sessions in 4 rats)
    • hippocampus, all pairs, REM: Wilcoxon signed rank test, z = 0.04, p = 0.965 (n = 34 sessions in 3 rats)
    • hippocampus, pyramidal pairs, REM: Wilcoxon signed rank test, z = -0.67, p = 0.499 (n = 34 sessions in 3 rats)
    • BLA, all pairs, REM: Wilcoxon signed rank test, z = 0.84, p = 0.39948 (n = 28 sessions in 4 rats)
    • BLA, pyramidal pairs, REM: Wilcoxon signed rank test, z = 1.40, p = 0.158 (n = 26 sessions in 4 rats)

    The criterion needs four NREM tests, two on putative pyramidal pairs, so it cannot be decided. As a description, with all pairs the hippocampus reactivates in NREM (23 sessions, EV above REV in 19, p = 0.0054) and the BLA does not reach p < 0.05 (16 sessions, 9, p = 0.14); in REM neither does (p = 0.83 and 0.21). Since the BLA all-pairs NREM test fails, the claim could at best be partially reproduced even with putative types. A BLA-only variant on the identified excitatory units (five Rat08 sessions, each with at least six) gives EV above REV in 4 of 5, p = 0.44, and five sessions cannot reach p < 0.05 (smallest two-sided p 0.0625); no file has six identified excitatory hippocampal units. No outcome in the earlier attempt.

  • BlockedResults, A subset of BLA cells are modulated during hippocampal SPW-Rs, paragraph 1; Figure 4a· Metadata insufficient

    A subset of BLA neurons is positively modulated during hippocampal SPW-Rs, 25.8% of interneurons (42 of 163) and 11.1% of pyramidal cells (137 of 1,233).

    A fraction of BLA neurons were significantly and positively modulated (‘upmodulation’; 42 of 163 interneurons (25.8%), 137 of 1,233 pyramidal cells (11.1%)) ... during hippocampal SPW-Rs (Fig. 4).

    Results, A subset of BLA cells are modulated during hippocampal SPW-Rs, paragraph 1; Figure 4a

    The claim gives the fractions of putative BLA interneurons and pyramidal cells that are upmodulated, and the dandiset has no putative types. As a description, the paper's Poisson test (M6) on all 822 BLA units of 17 sessions finds 105 (12.8%) upmodulated under O2 reading a (110 under b); among the identified units of five files, 8 of 36 inhibitory (22.2%) and 9 of 152 excitatory (5.9%). The earlier attempt reproduced this id with another test (spikes in ripple windows against shifted windows, all BLA units) on the abstract's statement; the claim now states fractions by cell type, which the dandiset cannot give.

  • BlockedResults, A subset of BLA cells are modulated during hippocampal SPW-Rs, paragraph 1; Figure 4a· Metadata insufficient

    Another subset of BLA neurons is negatively modulated during hippocampal SPW-Rs, 14.7% of interneurons (24 of 163) and 8.3% of pyramidal cells (102 of 1,233).

    or negatively modulated (‘downmodulation’; 24 of 163 interneurons (14.7%), 102 of 1,233 pyramidal cells (8.3%)) during hippocampal SPW-Rs (Fig. 4)

    Results, A subset of BLA cells are modulated during hippocampal SPW-Rs, paragraph 1; Figure 4a

    Fractions by putative cell type, and depends on bla-ripple-up-subgroup. As a description, 85 of 822 BLA units (10.3%) are downmodulated under O2 reading a. No outcome in the earlier attempt.

  • BlockedResults, Ripple-modulated BLA cells are preferentially involved in reactivations, paragraph 1; Supplementary Table 2· Metadata insufficient

    During run, 6.69% of hippocampus-BLA pyramidal pairs (2,521 of 37,660) are significantly positively correlated, 3.34% (1,258) negatively, and 89.96% (33,881) not reliably correlated.

    During run, a fraction of hippocampal–BLA pyramidal neuron pairs showed significantly positively correlated spike trains (2,521 of 37,660; 6.69%). Another small percentage (1,258 of 37,660; 3.34%) was negatively correlated, while the remaining majority (33,881 of 37,660; 89.96%) was not reliably correlated

    Results, Ripple-modulated BLA cells are preferentially involved in reactivations, paragraph 1; Supplementary Table 2

    Fractions of hippocampus-BLA putative pyramidal pairs, which the dandiset cannot select. No outcome in the earlier attempt.

  • BlockedResults, Ripple-modulated BLA cells are preferentially involved in reactivations, paragraph 1; Figure 5a; Supplementary Fig. 6a; Supplementary Table 2· Metadata insufficient

    Among nine groups of hippocampus-BLA pyramidal pairs formed by run correlation and BLA ripple modulation, the pairs with a significant positive run correlation and an upmodulated BLA partner show the largest pre-NREM to post-NREM increase in correlation, and differ from every other group.

    Hippocampus–BLA pairs with significant, positive run correlations and SPW-R upmodulation of the BLA partner showed the largest pre-NREM to post-NREM change

    Results, Ripple-modulated BLA cells are preferentially involved in reactivations, paragraph 1; Figure 5a; Supplementary Fig. 6a; Supplementary Table 2

    What the paper reports

    • the nine groups of pairs by run correlation and BLA ripple modulation: one-way ANOVA, F = 29.09, d.f. = 8, p = 1.1 × 10⁻⁴⁵ (n = 37,660 pairs in 3 rats)
    • the upmodulated, positively correlated group against the other groups: post hoc comparisons, p < 0.001, Tukey-Kramer (Fig. 5a legend, which gives P < 0.01; the Online Methods say Bonferroni)-corrected

    The criterion crosses putative pyramidal pairs with the ripple modulation of their BLA partner, and the dandiset has no putative types. The earlier attempt reproduced this id with the weights of BLA units in singular-vector joint patterns, a measure the claim no longer uses; the claim now tests the nine pair groups of the paper's Fig. 5a, which cannot be formed here.

  • BlockedResults, Ripple-modulated BLA cells are preferentially involved in reactivations, paragraph 1; Supplementary Fig. 6b· Metadata insufficient

    EV computed within each of the nine pair groups confirms the dominant contribution of the pairs with positive run correlation and an upmodulated BLA partner.

    The dominant contribution of this specific subgroup of cell pairs to reactivations was confirmed by calculating the EVs across the nine subgroups

    Results, Ripple-modulated BLA cells are preferentially involved in reactivations, paragraph 1; Supplementary Fig. 6b

    Depends on reactivation-involves-ripple-up-cells (pyramidal pair groups). No outcome in the earlier attempt.

  • BlockedResults, Ripple-modulated BLA cells are preferentially involved in reactivations, paragraph 2; Figure 5b; Supplementary Fig. 8a· Metadata insufficient

    Upmodulated BLA cells in the quartile contributing most to the reactivation increase their SPW-R gain from pre-NREM to post-NREM, while upmodulated cells of the other three quartiles do not.

    We found that upmodulated BLA neurons in the most strongly contributing quartile showed a specific increase in SPW-R gain (that is, firing rate during versus outside SPW-R) from pre-NREM to post-NREM compared to the upmodulated cells of the remaining three, low-contribution quartiles

    Results, Ripple-modulated BLA cells are preferentially involved in reactivations, paragraph 2; Figure 5b; Supplementary Fig. 8a

    What the paper reports

    • post-NREM against pre-NREM gain, upmodulated cells of the highest-contribution quartile: Wilcoxon one-tail signed rank test, z = -3.913, p = 4.6 × 10⁻⁵ (n = 63 cells in 3 rats)
    • the same, upmodulated cells of the remaining three quartiles: Wilcoxon one-tail signed rank test, z = 2.408, p = 0.992 (n = 73 cells in 3 rats)
    • second quartile, upmodulated cells (Supplementary Fig. 8a): Wilcoxon signed rank one-tailed test, z = 0.7839, p = 0.7834 (n = 18 cells)
    • third quartile, upmodulated cells (Supplementary Fig. 8a): Wilcoxon signed rank one-tailed test, z = 1.8873, p = 0.9704 (n = 16 cells)
    • fourth quartile, upmodulated cells (Supplementary Fig. 8a): Wilcoxon signed rank one-tailed test, z = 1.5420, p = 0.9384 (n = 39 cells)

    Contributions are computed over pooled putative pyramidal pairs, and the cells are upmodulated by bla-ripple-up-subgroup, both of which need putative types. No outcome in the earlier attempt.

  • BlockedResults, Ripple-modulated BLA cells are preferentially involved in reactivations, paragraph 2; Supplementary Fig. 9· Metadata insufficient

    Hippocampus-BLA EV does not depend on firing rates; EV exceeds REV across firing-rate octiles of the pairs, of the BLA cells and of the hippocampal cells.

    This control showed that EV did not depend on firing rates (Supplementary Fig. 9).

    Results, Ripple-modulated BLA cells are preferentially involved in reactivations, paragraph 2; Supplementary Fig. 9

    Uses the pooled pyramidal pairs of run-correlated-pair-fractions. No outcome in the earlier attempt.

  • BlockedFigure 6a, top; Results, The aversive trajectory is reactivated during SPW-Rs, paragraph 1· Metadata insufficient

    The hippocampus-BLA correlation pattern of the air puff trajectory is reactivated more around post-NREM ripples than around pre-NREM ripples.

    reactivation strength in a 500-ms window around ripple peaks (gray bars) was significantly higher in post-NREM compared to pre-NREM for air puff trajectory

    Figure 6a, top; Results, The aversive trajectory is reactivated during SPW-Rs, paragraph 1

    What the paper reports

    • post-NREM against pre-NREM ripple-peak reactivation strength, air puff trajectory, pyramidal pairs: Wilcoxon one-tail signed rank test, z = -3.793, p = 7.4 × 10⁻⁵ (n = 25 sessions in 3 rats)

    Pyramidal cells. As a description, the all-cell air puff template gives post above pre in 10 of 14 sessions, one-tailed p = 0.01 (under direction-all-cell-pairs). No outcome in the earlier attempt.

  • BlockedFigure 6a, bottom; Results, The aversive trajectory is reactivated during SPW-Rs, paragraph 1· Metadata insufficient

    The hippocampus-BLA correlation pattern of the safe trajectory is not significantly more reactivated around post-NREM ripples than around pre-NREM ripples (pyramidal pairs).

    significantly higher in post-NREM compared to pre-NREM for air puff trajectory ... but not for safe trajectory

    Figure 6a, bottom; Results, The aversive trajectory is reactivated during SPW-Rs, paragraph 1

    What the paper reports

    • post-NREM against pre-NREM ripple-peak reactivation strength, safe trajectory, pyramidal pairs: Wilcoxon one-tail signed rank test, z = -0.780, p = 0.217 (n = 25 sessions in 3 rats)

    Pyramidal cells. As a description, the all-cell safe template gives post above pre in 12 of 14 sessions, one-tailed p = 0.0043. No outcome in the earlier attempt.

  • BlockedFigure 6b; Results, The aversive trajectory is reactivated during SPW-Rs, paragraph 1· Metadata insufficient

    The post-NREM minus pre-NREM change in reactivation strength at the ripple peak is larger for the air puff trajectory than for the safe trajectory (pyramidal pairs).

    The pre-NREM vs. post-NREM difference in reactivation strength (RS) at the ripple peak is significantly higher for the air puff trajectory compared to safe trajectory

    Figure 6b; Results, The aversive trajectory is reactivated during SPW-Rs, paragraph 1

    What the paper reports

    • post-NREM minus pre-NREM ripple-peak reactivation strength, air puff against safe trajectory: Wilcoxon one-tail signed rank-test, z = -2.798, p = 0.00257 (n = 25 sessions)

    The criterion compares the post minus pre ripple-peak strength R of air puff and safe templates of pyramidal cells. The earlier attempt found this id not reproduced (p = 0.11) with singular-vector patterns over all units, a measure the paper does not use; the claim now uses R on pyramidal cells, which the dandiset cannot select. As a description, with all cells the air puff change does not exceed the safe one (8 of 14 sessions, one-tailed p = 0.48).

  • BlockedResults, The aversive trajectory is reactivated during SPW-Rs, paragraph 1· Too large to analyse

    Firing rates of BLA cells do not differ significantly between the safe and the air puff trajectories.

    Because the firing rates of BLA cells did not significantly differ between safe and air puff trajectories

    Results, The aversive trajectory is reactivated during SPW-Rs, paragraph 1

    What the paper reports

    • BLA pyramidal cells: Wilcoxon signed rank test, z = 1.101, p = 0.27
    • all BLA cells: Wilcoxon signed rank test, z = 0.302, p = 0.763

    The criterion gives not reproduced when either test (BLA pyramidal cells, all BLA cells) has p < 0.05 under every reading, and its readings include O7 reading b, the mean of each cell's position firing-rate curve, which needs position that only tracking the raw video would give (too large, as for rats-learn-puff-location). Under the readings that can be run (O7 reading a, rates over laps from reward times), the all-cell test has p < 0.05 under both session readings, median air puff minus safe 0.0036 Hz, p = 0.022 (739 cells, 14 sessions) and 0.0032 Hz, p = 0.0048 (1033 cells, 21 sessions), the direction of not reproduced; only O7 reading b keeps the claim from that outcome. The pyramidal test cannot be run (no putative types) and is not needed for not reproduced. The difference may come from the platform time and stops these laps hold, not from the trajectory. With laps of at most 30 s it remains (0.0050 Hz, p = 0.016, 466 cells in 11 sessions; 0.0030 Hz, p = 0.0043, 760 cells in 18 sessions), but with each lap cut to its last 10 s it is gone (median 0.0000 Hz, p = 0.75 and 0.82); these variants are the author's and decide nothing. Blocked as metadata_insufficient before the verifier's round 1, which pointed out that the all-cell test alone can decide. No outcome in the earlier attempt.

  • BlockedResults, NREM contributes to reinstatement of new place-threat representations, paragraph 1; Figure 7b; Supplementary Fig. 7c· Metadata insufficient

    For the 2.5% most strongly contributing hippocampus-BLA pairs, the pre-run to post-run change in correlation correlates with the pre-NREM to post-NREM change (r = 0.27), but not for the 2.5% least contributing pairs.

    We found that for strongly contributing, but not for weakly contributing, pairs the increase between pre-run and post-run coactivity was significantly correlated with the increase in coactivity between pre-NREM and post-NREM (Fig. 7b).

    Results, NREM contributes to reinstatement of new place-threat representations, paragraph 1; Figure 7b; Supplementary Fig. 7c

    What the paper reports

    • strongly contributing pairs (2.5% highest contributions): Pearson correlation, r = 0.27, p = 2.1 × 10⁻¹⁶ (n = 866 pairs, 3 rats)
    • weakly contributing pairs (2.5% lowest contributions): Pearson correlation, r = 0.03, p = 0.305 (n = 845 pairs, 3 rats)

    Depends on contributing-up-cells-gain-increase (pair contributions over putative pyramidal pairs). No outcome in the earlier attempt.

  • BlockedResults, NREM contributes to reinstatement of new place-threat representations, paragraph 1; Figure 7c; Supplementary Fig. 7b· Metadata insufficient

    The coupling between run and sleep correlation changes holds when the most and least contributing quartiles of pairs are compared (r = 0.13 against r = 0.013).

    This result was maintained when the most strongly and the most weakly contributing quartiles (instead of 2.5th percentiles) of the distribution were compared (Fig. 7c).

    Results, NREM contributes to reinstatement of new place-threat representations, paragraph 1; Figure 7c; Supplementary Fig. 7b

    What the paper reports

    • most strongly contributing quartile: Pearson correlation, r = 0.13, p = 1.1 × 10⁻³² (n = 9,415 pairs, 3 rats)
    • least contributing quartile: Pearson correlation, r = 0.013, p = 0.22 (n = 9,415 pairs, 3 rats)

    Depends on contributing-pairs-run-sleep-coupling. No outcome in the earlier attempt.

Findings about the data

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

  • Description missingNew in this attempt. The putative cell types the paper uses (pyramidal cell or interneuron for every unit, from monosynaptic connections and then k-means on the waveform) are not in the dandiset, and there are no spike waveforms. Five of the 33 files with units (Rat08-20130709 to 20130713) have a cell_type column with the values excitatory, inhibitory and unclassified, described only as "Type of cell this has been classified as."; it marks 152 excitatory and 36 inhibitory BLA units and 9 excitatory hippocampal units, consistent with the paper's monosynaptic identification alone. A short-latency cross-correlogram check supports this reading of the column. At p < 1e-6, 99.3% of the excitatory BLA units excite and 94.4% of the inhibitory ones inhibit another loaded unit at 1 to 4 ms, against 44% and 7.9% of the unclassified ones. Most of the paper's reactivation analyses cannot be repeated (17 claims blocked for this alone). The earlier attempt did not read this column.
  • Data missingNew in this attempt. The dandiset holds 40 recording days with 4702 units, where the paper reports 61 sessions with 7,390 units; per rat it holds 2317, 515, 440 and 1430 units (Rat08 to Rat11) against 2,444, 1,294, 1,138 and 2,514. Which rat is which of the paper's rats 1 to 4 is not stated.
  • Metadata errorNew in this attempt. The archive's asset metadata (variableMeasured) is wrong for all 40 files on two entries. It lists ElectricalSeries for the 24 files that hold no raw voltage and not for the 16 that hold it (acquisition/ElectricalSeries in Rat08-20130708 to 20130713 and all Rat09 files), and ProcessingModule for the 16 files without a behavior module and not for the 24 with one. Its Units entries agree with the files.
  • Metadata errorStill present (found by 000061-girardeau-2017-01; the version and the files are unchanged, and the draft holds the same assets). The epochs table is stored in milliseconds although its columns say seconds; in all 40 files its last stop time is exactly 1000 times the LFP duration.
  • Metadata errorStill present (found by 000061-girardeau-2017-01). Video timestamps are stated in seconds but are in milliseconds and are not on the session clock; the first timestamp matches the start of the video's epoch in 1 of 114 videos, while the spans match the epoch durations in 93 of 114.
  • File format errorStill present (found by 000061-girardeau-2017-01). In ten files the spike_times_index is wrong at the first units; unit 1's spike times fall back once, after 65536 times k spikes, and unit 0 stops where unit 1 starts, so the first part of unit 1 is the end of unit 0. Units 0 and 1 of these files were left out of the analyses.
  • Data missingStill present (found by 000061-girardeau-2017-01). Seven files have no units table, sixteen have no sleep states and fourteen have no ripple events; the Rat09 files have no hippocampal units and no ripples.
  • Data missingStill present (found by 000061-girardeau-2017-01). There is no position tracking and no air puff location, only the puff times and the raw video frames, so behaviour on the track cannot be measured without tracking about 1.9 million frames; laps and the air puff direction can only be inferred from reward and puff times.
  • Description missingStill present in the dandiset (found by 000061-girardeau-2017-01). Unit region labels are undescribed short codes (bla, blv, bmp, ladl, cen, den, ven, pir, plc01, PLCo1, cpu, hpc, HIP, no label), some in two cases, Rat09-20140328 has no location column, and the electrode table gives every channel as unknown. The paper's Supplementary Fig. 3 names the structures the codes match (BLA, BLV, LaDL, BMA, BMP, Pir, VEn, DEn), which lets this attempt read them; the dandiset still does not describe them.
  • Description missingStill present (found by 000061-girardeau-2017-01). The trials table of the Rat08 and Rat09 files is described only as "experimental trials"; its intervals of 3 to 5 s fall partly outside the run and are not laps.
  • Description missingStill present (found by 000061-girardeau-2017-01). The ripple annotations carry an undescribed number in each label (for example "Ripple peak 156"), one per file, presumably the detection channel; ripples, puffs and rewards are stored as annotation series under stimulus presentation.
  • Methods not availablePartly fixed (found by 000061-girardeau-2017-01, when the paper's text was not available). The full text was supplied for this attempt and the paper links its code (github.com/buzsakilab/papers, GGirardeau-BLAHpcInteractions-Package), which the claims file's reader read; the dandiset itself still links no analysis code, its related resources naming only the paper.

Notebooks

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

Reproduction Hippocampus and amygdala reactivation during sleep (Girardeau et al. 2017)

Tests the claims of Girardeau, Inema and Buzsaki (2017), written from the full text, the supplement and the paper's code, on dandiset 000061, which holds sorted units of the hippocampus and the amygdala, sleep states, ripple, air puff and reward events from four rats on an air puff linear track task. It tests the unit counts, habituation over days, BLA firing by sleep state, hippocampus-BLA and hippocampus-central nuclei reactivation (explained variance) with all cells, and the ripple rate before and after training. Claims on putative pyramidal cells and on position are blocked, as the dandiset holds neither. All files

Verification

  • Round 1
    2026-10-20    Agreed with 32 of 32 outcomes. Independent agent run by CatalystNeuro (same harness, separate context, did not write the notebook). Reviewed: Round 1. The executed notebook with the conclusion cells and the summary removed, the helper module and the claims file, with the paper. The verifier queried the archive's asset metadata and tested helper functions on synthetic inputs (EV and REV, pair correlations, the Poisson tails, the peri-ripple bins, the interval helpers), finding them correct. Its report had no attempt record and gave its own outcome for each claim; all 32 agree with the record, and for bla-rate-same-across-directions it gave another blocking reason. Points not tied to one claim: the raw voltage reported in 16 files disagrees with the asset metadata (24 other files); the author printed the groups of every file, which confirm the 16 files, and recorded the asset metadata as wrong for ElectricalSeries and ProcessingModule in all 40 files (new finding). M3's rule dropping spikes after the last run epoch is not implemented; the author kept it so and says in the helper docstring why it cannot change any bin used.
  • Round 2
    2026-10-21    Agreed with 32 of 32 outcomes. Independent agent run by CatalystNeuro (same harness, separate context, did not write the notebook). Reviewed: Round 2. The full executed notebook with its conclusions and summary, the helper module, the claims file, the attempt record and the README, with the paper. The verifier agreed with all 32 outcomes and found no place where the claims file misreads the paper. Points not tied to one claim. The README's run time (10.0 min, six of them streaming) did not match cell 77 (7.2 min) and cell 1 (302 s of loading); the README now gives the final run of this revision (10.2 min, 366 s of loading). The round-1 reasons were found faithful, except that the platform-time point was attributed to habituation-over-days; that entry is corrected.

Method differences

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

  • BLA is the bla label (case ignored); units 0 and 1 of the files with a wrong spike index are left out.ripple-modulated-cells-larger-reactivation reactivation-peaks-at-ripples direction-all-cell-pairs
  • The dandiset does not say which of Rat08 to Rat11 are the paper's rats 1 to 4. Rat09, the rat without hippocampal units, is taken as rat 2 (whose hippocampal probe failed) and the others in order; this is an assumption.unit-counts
  • Units are counted as stored, one per session, including units 0 and 1 of the ten files whose spike index is wrong; location counts use the dandiset's labels read as the paper's structures.unit-counts
  • Laps are O1 reading c, the intervals between consecutive rewards at opposite ends with both rewards inside the run. Reading a has no lap intervals in the dandiset (its trials table is not laps) and reading b needs position, which it lacks. As answered, the claim is judged on reading c alone; the hand-picked 150-cm window of ClassifyLaps.m is not reproduced by any reading. A reading c lap runs from reward to reward, so it includes the time on the starting platform and any stop, which the M8 laps leave out; the count of laps is the same either way.habituation-over-days
  • The air puff direction of a day is the lap direction holding most puffs (1153 of 1197 puffs fall in run laps of that direction); the dandiset gives no direction. Rat08-20130709 has no puff inside a run lap and has no air puff count.habituation-over-days
  • Training day is the rank of the recording date within each rat, over the 40 files; the paper's 55 behaviour sessions include days the dandiset does not hold.habituation-over-days
  • Cells are the units the dandiset's cell_type column marks excitatory, read as the paper's monosynaptically identified pyramidal cells (O4 reading a). The column exists in five Rat08 files only, so the test uses one rat and about a fifth of the paper's cells. Reading b (all putative pyramidal cells) cannot be formed; as answered, it is not needed for the outcome.bla-pyr-sleep-firing-increase
  • Units 0 and 1 of files with a wrong spike index are left out (Rat08-20130710 and 20130711).bla-pyr-sleep-firing-increase
  • Cells are the units marked inhibitory in the five typed Rat08 files, read as the paper's identified interneurons (O4 reading a, which as answered decides alone); reading b cannot be formed.bla-int-sleep-firing
  • BLA is the bla label (case ignored); the wider amygdala labels are a variant.reactivation-all-cell-pairs
  • Units 0 and 1 of the files with a wrong spike index are left out.reactivation-all-cell-pairs
  • A 50-ms bin belongs to an epoch when its centre does, a standard choice.reactivation-all-cell-pairs
  • Central nuclei are the cen label. Units 0 and 1 of files with a wrong spike index are left out.cen-no-reactivation
  • Only the all-cell test is run; the pyramidal-pair test needs putative types the dandiset lacks.ripple-modulated-cells-larger-reactivation
  • Only the all-cell test is run; the pyramidal test needs putative types the dandiset lacks.reactivation-peaks-at-ripples
  • Laps are O1 reading c (between rewards at opposite ends); the dandiset has no lap intervals or position. The air puff direction of a day is the lap direction holding most puffs. These laps include the time on the starting platform after each reward and any stop, so the air puff and safe templates hold awake activity off the track that the M8 laps (position inside a 150-cm window) leave out.direction-all-cell-pairs
  • Ripples are the dandiset's rip annotations, counted in NREM when their peak is, a standard choice.ripple-rate-unchanged

Provenance

Attempt record
000061-girardeau-2017-02
Paper
Reactivations of emotional memory in the hippocampus–amygdala system during sleep
Date
2026-10-20
Made by
An agent, run by CatalystNeuro
Dataset
DANDI:000061
Version
0.250624.0426 (doi:10.48324/dandi.000061/0.250624.0426)
Run on
not recorded
Harness
Claude Code (GitHub Actions)
Model
claude-opus-5-5
Protocol
pilot-10

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.

Earlier attempts: 000061-girardeau-2017-01 (2026-10-16)

← DANDI:000059DANDI:000067 →