Maintenance of persistent activity in a frontal thalamocortical loop (DANDI:000009)

Claims

From the paper Maintenance of persistent activity in a frontal thalamocortical loop

Reproducing Guo et al. (2017) from dandiset 000009 27

Notebook reproduce_guo_2017.ipynb

Not reproducedResults, Thalamus is required for motor preparation, paragraph 1verifier disagreed at first
Mice performed the delayed-response task at 77.6% correct on average, with 10.3% lick-early trials and 1.1% no-response trials.
mean percent correct 77.6%; lick-early before ‘go’ cue 10.3%, no response 1.1%
Results, Thalamus is required for motor preparation, paragraph 1
Over all 173 behavioural recordings percent correct is 70.66% (reading a, pooled) to 75.46% (reading c, mean over sessions), within 77.6 +/- 7.76%. Lick-early trials are 14.66 to 18.18% against 10.3 +/- 1.03%, and no-response trials 0.73 to 0.98% against 1.1 +/- 0.11%. No reading puts all three within tolerance. The 36 whole-cell recordings alone give 77.42 to 83.47% correct, 7.93 to 9.65% lick early and 0.96 to 1.55% no response, at most two values within tolerance. Per mouse, lick-early rates range from 3.6 to 37.2%. Partially reproduced before the first verification; changed to not reproduced because the criterion asks all three values under one reading and two of the three are out under every reading of the full set of recordings.
How the test differed from the paper’s method
- The dandiset holds recording sessions only; the paper's averages may come from behaviour sessions that are not deposited.
- Rows with no outcome and no pole times (5369 extracellular rows, type 'non-performing') are not counted as trials.
Not reproducedMethods, Electrophysiology data analysis, paragraph 2; Extended Data Figure 4bverifier disagreed at first
Of 1214 ALM units, 1006 were putative pyramidal neurons (spike width > 0.6 ms), 166 putative fast-spiking (< 0.4 ms) and 42 unclassified; the spike-width distribution was bimodal.
In ALM, 1214 single units were recorded across 57 behavioral sessions. ... units with width < 0.4 ms were defined as putative fast-spiking (FS) neurons (166/1214) and units with width > 0.6 ms as putative pyramidal neurons (1006/1214)
Methods, Electrophysiology data analysis, paragraph 2; Extended Data Figure 4b
The dandiset holds 966 ALM units in 47 sessions (paper 1214 in 57). By computed width (P1) 577 units exceed 0.6 ms and 127 are below 0.4 ms; by file label (P2) 810 are 'Wide width spiking' and 121 'Narrow width spiking'. The fractions under P2, 0.839 and 0.125, are close to the paper's 0.829 and 0.137; no count is within tolerance under either reading. 514 of the 577 units above 0.6 ms have their waveform peak at the last sample, so their widths are lower bounds. The width histogram has two modes (0.28 ms and 0.68 to 0.72 ms) with a trough at 0.36 ms (4 units). The answer of 2026-10-20 makes P1 the Methods' rule and keeps P2 as a reading; no count is within tolerance under either, so the outcome is unchanged, the fractions under P2 matching.
How the test differed from the paper’s method
- The deposited mean waveforms are too short to measure the widths the Methods use; both the computed widths (P1) and the files' labels (P2) are shown, as convention C says.
Not reproducedResults, Preparatory activity in thalamus, paragraph 1; Figure 2everifier disagreed at first
Most ALM putative pyramidal neurons (70%, 704/1006) distinguished trial types; 145 had preparatory activity only, 272 peri-movement selectivity only and 287 both.
For a majority of neurons, activity differed across trial types (70%, 704/1006; P < 0.05, t -test; Methods). ... subsets of neurons showed selective preparatory activity ( Fig. 2b , left column; Fig. 2e , 145/1006), selective peri-movement activity during the response epoch ( Fig. 2b , right column; Fig. 2e , 272/1006), or both ( Fig. 2b , middle column; Fig. 2e , 287/1006)
Results, Preparatory activity in thalamus, paragraph 1; Figure 2e
Under P1 and the first selectivity reading, 398 of 577 units are selective (95, 154, 149 by type). Under P2, 571 of 810 (70.5%) are selective against 70% in the paper, with 132 preparatory only, 222 peri-movement only and 217 both; scaled to 1006 units, 164, 276 and 270 against 145, 272 and 287. Over all 32 readings the selective fraction is 0.62 to 0.73. No reading puts the four counts within tolerance because the dandiset holds 577 or 810 pyramidal units against 1006; the fractions agree with the paper.
How the test differed from the paper’s method
- Each unit's trials are those from the trial holding its first spike to the trial holding its last, as the files give no observation interval per unit; this applies to every unit analysis.
- A window that ends after the last stored spike of a trial (spikes are stored to about 4.8 s after pole onset) is counted as missing for that trial; this applies to every unit analysis and touches 221 of 8991 kept ALM trials, 173 of 14997 thalamic and 19 of 3573 SNr for R1.
- Unit classes under readings P1 and P2 of convention C (answer of 2026-10-20; P3 does not apply to ALM). Outcome unchanged.
Not reproducedMethods, Electrophysiology data analysis, paragraph 3; Results, Preparatory activity in thalamus, paragraph 2; Extended Data Figure 4everifier disagreed at first
Of 790 thalamic units, 672 were putative thalamic neurons (spike width > 0.5 ms) and 313 lay within 0.4 mm of the VM/VAL centre; 295 VM/VAL units were analysed.
In thalamus, 790 single units were recorded across 73 behavioral sessions. ... Units with spike width > 0.5 ms were selected as putative thalamic neurons (672/790) ... units within 0.4 mm from the VM/VAL center ... were scored as VM/VAL neurons (313/790)
Methods, Electrophysiology data analysis, paragraph 3; Results, Preparatory activity in thalamus, paragraph 2; Extended Data Figure 4e
The dandiset holds 789 thalamic units in 72 sessions (paper 790 in 73). By file label (P2) 682 are putative thalamic units, within 672 +/- 67; by computed width (P1) only 148. 259 units lie within 0.4 mm of the L2 centre (313 +/- 31) and 222 (P2) or 20 (P1) putative units among them (295 +/- 30). The total and the P2 class count are within tolerance, the location counts are not.
How the test differed from the paper’s method
- The files have no VM/VAL label (reading L1 cannot be applied); reading L2 uses the claims file's proxy centre.
- Unit classes under readings P1 and P2 of convention C; the criterion excludes P3 for these counts (answer of 2026-10-20). Outcome unchanged.
Not reproducedResults, Preparatory activity in thalamus, paragraph 2; Figure 2fverifier disagreed at first
A majority of VM/VAL neurons (69%, 204/295) distinguished trial types, a fraction indistinguishable from ALM; 43 had preparatory activity only, 82 peri-movement selectivity only and 79 both, a distribution similar to ALM.
A majority of VM/VAL neurons discriminated trial types (69%, 204/295; indistinguishable from ALM, chi-squared test, P = 0.79). ... VM/VAL neurons exhibited similar selectivity and time-course to ALM (chi-squared test, P > 0.5)
Results, Preparatory activity in thalamus, paragraph 2; Figure 2f
What the paper reports
- fraction selective, VM/VAL against ALM: chi-squared test, p = 0.79 (n = 295 VM/VAL and 1006 ALM neurons)
- selectivity types, VM/VAL against ALM: chi-squared test, p > 0.5
Under P1 only 20 VM/VAL units qualify (13 selective under the first reading), and their types do not differ from ALM. Under P2, 149 of 222 (67%) are selective against 204/295 (69%), and the fraction does not differ from ALM's 70.5% (chi-squared p = 0.37). The types differ from ALM in all 16 P2 readings (2 x 4 and 2 x 3 tests p < 0.05; first reading 38, 75 and 36, p = 0.011 and 0.0053), where the paper finds p > 0.5. The answer of 2026-10-20 adds reading P3 (all 259 VM/VAL units regardless of width). Under P3, 174 are selective under the first reading (156 to 177 over the readings, below 204 +/- 20), the fraction does not differ from ALM in any reading, and the types differ from ALM in all 32 readings (first reading 40, 89 and 45; p = 0.024 and 0.0098 against ALM P1). Counts are below tolerance in every reading. The selective fraction holds; the type distribution holds only on the 20 P1 units. Outcome unchanged.
How the test differed from the paper’s method
- VM/VAL by reading L2 only; unit classes under P1, P2 and P3 of convention C.
Not reproducedResults, Thalamus drives ALM preparatory activity, paragraph 1; Figure 3fverifier disagreed at first
Strong photoinhibition in VGAT-ChR2-EYFP mice abolished thalALM activity nearly completely, to 2% of control.
Photoinhibition abolished thal ALM activity nearly completely (to 2 % of control without photoinhibition) ( Fig. 3f )
Results, Thalamus drives ALM preparatory activity, paragraph 1; Figure 3f
Putative thalamic units of the VGAT thalamic sessions fall to 1.1% (T1) or 0.9% (T2) of control (Q1; Q2 0.9 to 1.4%) on 93 (P1) or 113 (P2) units against 148 +/- 15; none is activated and 101 of 113 are significantly inhibited (P2, T1, Student). With all 134 units the count is within tolerance but narrow units raise Q1 to 18%. The silencing is nearly complete as the paper states; the 2% and the unit count are not both within tolerance under any reading. No VGAT thalamic unit lies within 0.4 mm of the L2 centre, so reading (b) has no units. One mouse gives Q1 1.8% and the other 35.2% with all units.
How the test differed from the paper’s method
- The files give the photostimulus power as 2.23 mW; one session was photostimulated in the sample epoch and drops out.
- Unit classes under P1 and P2 of convention C (Fig. 3, so P3 does not apply; all units are shown as a variant). Outcome unchanged.
Not reproducedResults, Thalamus drives ALM preparatory activity, paragraph 2
During the delay epoch the mean membrane potential of ALM neurons was close to spike threshold, 12.7 mV below it on average.
the average membrane potential of ALM neurons was close to the spike threshold during the delay epoch (action potential threshold - mean basal membrane potential = 12.7 ± 0.7 mV; mean ± SEM, n = 60 cells)
Results, Thalamus drives ALM preparatory activity, paragraph 2
In the 36 deposited cells (paper 60), threshold minus basal Vm is 7.19 +/- 0.43 mV with the Methods' baseline (a) and 6.46 +/- 0.35 mV with the delay epoch (b), against 12.7 +/- 0.7 mV. With the deposited kink, 6.88 and 6.15 mV. Leaving out cell_21_1, whose trial start times are shifted by one row, changes these by at most 0.14 mV. The kink is the first sample searched (1.5 ms before the peak) in 1.7% of spikes (0.4 to 6.8% per cell); leaving those spikes out gives 7.23 mV with baseline (a).
How the test differed from the paper’s method
- The threshold is the Vm at the kink of convention V, averaged over every spike of the recording.
- Spike peaks are located from the deposited spike-removed trace; the kink and clipping are computed by the claims file's rule and match the deposited trace.
Not reproducedResults, Thalamus drives ALM preparatory activity, paragraphs 2 and 3; Figure 3g, 4b, cverifier disagreed at first
Thalamus photoinhibition hyperpolarized ALM neurons by 18.7 mV, preventing spiking, with an onset 7.9 ms after photostimulus onset.
Thal ALM photoinhibition caused hyperpolarization (Membrane potential, control trials –photoinhibition trials, -18.7 ± 1.1 mV; measured for 100 ms, starting 20 ms after photostimulus onset; mean ± SEM, n = 16 cells), preventing ALM neurons from spiking
Results, Thalamus drives ALM preparatory activity, paragraphs 2 and 3; Figure 3g, 4b, c
The criterion's onset readings are E (1.3 s after pole_in_time) and M (0.1 s after pole_out_time); the files also hold the laser trace, an OptogeneticSeries the first attempt did not read, whose onset (reading L, asked about in question 7) lies 1.4725 s after pole_in_time in these cells, 28 ms before M. Under E the 20 to 120 ms window precedes the change and Vm changes by +0.31 mV. Under L thalamus photoinhibition changes Vm by -13.68 +/- 1.14 mV (all 16 cells hyperpolarize, -4.7 to -21.7 mV) and the spike rate falls from 3.24 to 0.14 spikes/s; the onset is 9.60 +/- 1.98 ms by V2, within 7.9 +/- 1.7, and 0.40 ms by V1, which is not. Under M the change is -16.83 +/- 1.24 mV and the rate falls from 3.14 to 0.06 spikes/s; the onset (9.05 ms by V1 under M) is not judged, as no reading the claims file lists gives the laser time. The hyperpolarization and the loss of spiking are present at 73% (L) or 90% (M) of the paper's -18.7 +/- 1.1 mV; no reading puts both values within tolerance. Measured is reading M, a reading the criterion lists.
How the test differed from the paper’s method
- The photostimulus onset is placed by readings E and M of the criterion, and by the onset of the deposited laser trace (L), a reading not in the claims file, shown pending question 7.
- The criterion has no partial clause; the hyperpolarization and the silencing are present under M and L at 73 to 90% of the paper's value, but no listed reading meets the criterion.
Not reproducedResults, Thalamus drives ALM preparatory activity, paragraph 3; Figure 4d-fverifier disagreed at first
Photoinhibition of M1 during the delay epoch caused only a slight hyperpolarization (3.3 mV) and a small drop in spike rate (1.4 spikes per s) in ALM neurons.
direct photoinhibition of M1 during the delay epoch ... caused only a slight hyperpolarization of the membrane potential and spike rates of ALM neurons (-3.3 ± 1.0 mV; -1.4 ± 0.6 spikes per s; mean ± SEM, n = 11 cells)
Results, Thalamus drives ALM preparatory activity, paragraph 3; Figure 4d-f
Under M the 11 cells change by -2.75 +/- 1.30 mV, within -3.3 +/- 1.0, and by +0.76 +/- 1.29 spikes/s, outside -1.4 +/- 0.6. Under the laser onset (L) they change by -3.26 +/- 1.08 mV, within, and by -0.01 +/- 1.16 spikes/s, outside; under E by -0.09 mV and +0.94 spikes/s, neither within. The slight hyperpolarization holds under M and L; the drop in spike rate is not seen under any reading.
How the test differed from the paper’s method
- The photostimulus onset is placed by readings E and M, and by the laser onset (L, question 7).
- The criterion has no partial clause; one of its two values is within tolerance under M and L.
Not reproducedResults, ALM drives thalamic preparatory activity, paragraph 1; Figure 6c
ALM photoinhibition decreased the activity of VM/VAL neurons to 25% of control, inhibiting 190 of 201 neurons (139 significantly) and significantly activating none.
ALM photoinhibition decreased activity in a majority of VM/VAL neurons (to 25% of control; 190/201 inhibited, 11/201 activated, 139/201 significantly inhibited, 0/201 significantly activated, P < 0.05, t -test, Fig. 6c )
Results, ALM drives thalamic preparatory activity, paragraph 1; Figure 6c
Under P1 only 15 VM/VAL units qualify (to 50.0% of control). Under P2, 172 units from 7 mice fall to 47.5% (Q1), 46.5% (Q2) or 36.4% (Q3) of control under ALM photoinhibition at 1.5 mW against 25%; 151 are inhibited, 21 activated, 99 (Student) or 105 (Welch) significantly inhibited and 2 significantly activated. The answers of 2026-10-20 fix the power at 1.5 mW (as already run) and add reading P3 (all VM/VAL units). Under P3 with T1, 201 units qualify, the paper's n, and 173 are inhibited (within 190 +/- 19), but activity falls only to 55.9% (Q1) or 40.0% (Q3), 28 are activated, 106 (Student) or 117 (Welch) significantly inhibited and 4 or 5 significantly activated. No reading puts every value within tolerance, and the partial condition fails because 1 to 5 units are significantly activated under P2 and P3. The reduction is present, about half as deep as the paper's; with all ALM powers pooled it is to 57.7% (P2, T1). Outcome unchanged.
How the test differed from the paper’s method
- VM/VAL by reading L2; ALM photoinhibition at 1.5 mW (convention P), other powers as a variant; unit classes P1, P2 and P3.
Not reproducedResults, ALM drives thalamic preparatory activity, paragraph 1; Methods, Electrophysiology data analysis, paragraph 3; Figure 6d; Extended Data Figure 8b, cverifier disagreed at first
The reduction of thalamic activity by ALM photoinhibition was mainly limited to VM/VAL; neurons within 1.0 mm of the VM/VAL centre were inhibited to 36% of control, and neurons farther away less.
neurons within 1.0 mm from VM/VAL center showed robust inhibition (to 36% of control activity during the first 100 ms inhibition
Results, ALM drives thalamic preparatory activity, paragraph 1; Methods, Electrophysiology data analysis, paragraph 3; Figure 6d; Extended Data Figure 8b, c
Under P1 only 29 units qualify, none at 1.0 mm or more. Under P2 with the 20 to 120 ms window, activity falls to 48% (Q1) within 0.5 mm, 62 to 65% at 0.5 to 1.0 mm and 81 to 85% at 1.0 mm or more, so the reduction is concentrated near the centre. Within 1.0 mm it falls to 52.6 to 53.4% (Q1) or 39.6 to 45.5% (Q3); only Q3 with correct trials (39.6%) is within 36 +/- 3.6%. Near the centre the group counts (240 to 286) are within 250 +/- 25 except 286 (u2, T1); farther away (110 to 132, 26 or 27) they are below the paper's 160 and 46 in every reading.
How the test differed from the paper’s method
- Distances from the L2 proxy centre in the coronal plane; unit classes under P1 and P2, and all units (u2), which is reading P3 of the answer of 2026-10-20; under u2 activity within 1.0 mm falls to 56.4 to 58.3% (Q1) or 44.2 to 48.2% (Q3) in the 20 to 120 ms window. Outcome unchanged.
Not reproducedFigure 6e; Extended Data Figure 6f; Extended Data Figure 10d
After ALM photoinhibition onset, ALM pyramidal neurons were silenced after 8.1 ms and VM/VAL activity reduced after 10.9 ms; SNr neurons changed later, after 15.2 ms.
SNr neurons were affected by ALM photoinhibition with a relatively long latency difference (15.2 ± 4.6 ms, mean ± SEM, P < 0.05, t -test), longer than for reducing thal ALM activity (10.9 ± 2.9 ms; Fig. 6e )
Figure 6e; Extended Data Figure 6f; Extended Data Figure 10d
What the paper reports
- SNr against VM/VAL latency: t-test, p < 0.05
Under P1 no ALM pyramidal unit of the PV-Cre x Ai32 ALM sessions qualifies (computed widths are 0.6 ms or less there). Under P2, 151 units (paper 256) give an ALM onset of 6 or 7 ms (8.1 +/- 1.2); VM/VAL onsets are 14 ms (T1) or 4 ms (T2) against 10.9 +/- 2.9, and SNr onsets 12 to 32 ms against 15.2 +/- 4.6 (P2 SNr units, now the 'Narrow width spiking' label, with 35 inhibited and 53 modulated). No reading puts all three within tolerance with SNr latest; measured is reading O1 with T1 and SNr reading (a).
How the test differed from the paper’s method
- SNr populations from snr-photoinhibition under P2, 5 trials, T1, Student.
- The answer of 2026-10-20 adds P3 for VM/VAL; its 201 units give onsets of 14 ms (T1) and 7 or 14 ms (T2), and no reading with them puts all three onsets within tolerance with SNr latest. Outcome unchanged.
Not reproducedResults, ALM drives thalamic preparatory activity, paragraph 2; Extended Data Figure 10cverifier disagreed at first
ALM photoinhibition changed only a small fraction of SNr GABAergic neurons (48/181 significantly inhibited, 23/181 activated), more decreasing than increasing, and reduced SNr activity overall by 3.6 spikes per s (8.3%), which argues against SNr mediating the thalamic reduction.
Photoinhibition of ALM changed only a relatively small fraction of SNr neurons (48/181 significantly inhibited; 23/181 significantly activated, P < 0.05, t -test,). Moreover, neurons that decreased their activity were more numerous than neurons that increased their activity (bootstrapping over neurons, P < 0.01, Methods). Overall, inhibiting ALM reduced SNr activity by 3.6 spikes per second (8.3% of control activity
Results, ALM drives thalamic preparatory activity, paragraph 2; Extended Data Figure 10c
What the paper reports
- more down-modulated than up-modulated neurons: bootstrapping over neurons, p < 0.01 (n = 181 putative GABAergic SNr neurons)
The dandiset holds 181 SNr units (paper 227), 171 under P1 and 146 under P2 (the 'Narrow width spiking' label; before the first verification P2 took every unit not labelled wide, 168). Under P1 (no minimum of trials, T1, Student) 45 units are significantly inhibited and 21 activated (paper 48 and 23), the reduction is 2.33 spikes/s (5.97%), and more decrease than increase (bootstrap p = 0.0009); under P1 and T1, 44 to 50 and 21, p = 0.0003 to 0.0019. Under P2 and T1, 35 to 40 and 18 (p = 0.0020 to 0.0105). The reduction is 2.25 to 2.44 spikes/s (5.8 to 6.3%) under P1 and 2.68 to 2.98 (6.5 to 7.2%) under P2, against 3.6 spikes/s (8.3%), out of tolerance in every reading. Under P1 the counts and the bootstrap test agree with the paper; the reduction is weaker.
How the test differed from the paper’s method
- The 46 wider SNr units of the paper are not deposited; the dandiset's 181 units appear to be the putative GABAergic ones, shown also as a set of their own.
Partially reproducedResults, Thalamus drives ALM preparatory activity, paragraph 1; Figure 3c
Strong photoinhibition of thalALM (VGAT-ChR2-EYFP) decreased ALM activity to 6% of control, inhibiting 309 of 314 neurons (155 significantly) and significantly activating none.
Thal ALM photoinhibition decreased ALM activity to 6% of control, reducing activity in nearly all neurons ( Fig. 3c , 309/314 inhibited, 3/314 activated, 155/314 significantly inhibited, 0/314 significantly activated, t -test, P < 0.05)
Results, Thalamus drives ALM preparatory activity, paragraph 1; Figure 3c
In 234 (P1) or 261 (P2) VGAT ALM putative pyramidal units from 3 mice, thalamus photoinhibition at 10 mW lowers activity to 2.1 to 2.5% (Q1) or 5.1 to 5.2% (Q2) of control, the median ratio being 0. Under P2 and T1, 256 of 261 units are inhibited, 2 activated, 127 (Student) or 208 (Welch) significantly inhibited and none significantly activated. The unit counts (230 to 261) are below 314 +/- 31 and no reading puts every value within tolerance; the criterion's partial condition holds in every reading. With the PV-Cre x Ai32 ALM sessions at 10 mW (a variant), activity falls to 21.7% (P2, T1) and 7 units are significantly activated. The answer of 2026-10-20 confirms VGAT-ChR2-EYFP mice at 10 mW and keeps the PV-Cre x Ai32 sessions as a variant, as the attempt already ran; P1 and P2 stay the unit readings. Outcome unchanged.
How the test differed from the paper’s method
- Only the VGAT-ChR2-EYFP sessions at 10 mW, as the criterion says; the PV-Cre x Ai32 ALM sessions with thalamic photostimulation are shown as a variant, not judged.
- Unit classes under P1 and P2 of convention C.
Partially reproducedResults, ALM drives thalamic preparatory activity, paragraph 1; Figure 6c bottom
The few VM/VAL neurons not changed by ALM photoinhibition were significantly less selective in the delay epoch, while neurons strongly reduced by it carried trial-type information; the change of rate and delay selectivity were negatively related (regression slope -0.73, r -0.33).
A small fraction of neurons in VM/VAL was not changed by ALM photoinhibition. These neurons also showed significantly less selectivity ( Fig. 6c , P = 0.002, t -test)
Results, ALM drives thalamic preparatory activity, paragraph 1; Figure 6c bottom
What the paper reports
- delay selectivity, significantly against non-significantly modulated neurons: t-test, p = 0.002
Under P1 (15 neurons, first reading) the slope is -0.195 and r -0.445, and the group difference is not significant (p = 0.077). Under P2 with T1, the 71 neurons not significantly changed are less selective in the delay epoch than the 101 changed ones (1.13 against 1.77 spikes/s, p = 0.0030; paper P = 0.002), and the change of rate and selectivity are negatively correlated (r = -0.278; slope -0.074 with selectivity on change, -1.048 with change on selectivity; paper -0.73 and -0.33). The answer of 2026-10-20 adds P3 (all VM/VAL units); under it the group difference and a negative r hold in 12 of 16 readings (89 against 112 neurons, p = 0.0031, T1, Student), with r -0.142 to -0.201. The group difference and a negative r hold in all 16 P2 readings and 4 of 16 P1 readings; slope and r are out of tolerance in every reading. Outcome unchanged.
How the test differed from the paper’s method
- Unit classes under P1, P2 and P3 of convention C.
ReproducedResults, Thalamus is required for motor preparation, paragraph 1; Figure 1c
Unilateral photoinhibition of ALM during the delay epoch biased responses to the ipsilateral side, lowering performance on contra trials and raising it on ipsi trials.
Unilateral photoinhibition of ALM during the delay epoch produced an ipsilateral response bias ... reducing performance for contralateral trials and increasing performance for ipsilateral trials ( P < 0.001; paired t -test, Fig. 1c )
Results, Thalamus is required for motor preparation, paragraph 1; Figure 1c
What the paper reports
- paired t-test, p < 0.001 (n = 84 behavioural sessions, 11 mice)
In 66 recording sessions of 11 PV-Cre x Ai32 mice, ALM photoinhibition at 1.5 mW in the delay epoch lowers contra performance from 0.774 to 0.368 (paired t = -16.67, p = 3.6e-25; no session increases) and raises ipsi performance from 0.673 to 0.744 (t = 3.99, p = 0.00017; 46 of 66 sessions increase). With all photoinhibition powers pooled, p = 4.7e-25 and 0.00054. Contra performance falls in every mouse. The answer of 2026-10-20 (convention P) puts the 0.75 and 0.79 mW trials outside the condition, as the main test already did; the outcome is unchanged.
How the test differed from the paper’s method
- Tested on the behavioural trials of the 66 recording sessions (ALM, thalamic and SNr recordings), not on the paper's 84 behaviour sessions, which are not in the dandiset.
- Photoinhibition trials at 1.5 mW (the Methods' power, convention P); trials at 0.75 or 0.79 mW are pooled in a variant only.
ReproducedMethods, Photoinhibition, paragraph 3; Extended Data Figure 2d
In VGAT-ChR2-EYFP mice, thalamus photoinhibition reduced activity 0.5 to 1.1 mm from the fibre tip to 15.9% of control.
Photoinhibition reduced activity (0.5-1.1 mm from the tip of the optical fiber) to 15.9 ± 9.3%
Methods, Photoinhibition, paragraph 3; Extended Data Figure 2d
Over the whole photostimulus (reading W1), putative thalamic units 0.5 to 1.1 mm from the fibre site fall to 17.0 to 20.3% of control (Q1) or 16.5 to 19.3% (Q3), within 15.9 +/- 9.3%; Q2 gives 37.5 to 51.2%. In the 20 to 120 ms window (W2) they fall to 1.2 to 1.5% (Q1). 63 to 79 units qualify (paper 114), grouped 23 to 24, 21 to 24 and 31 to 42 near 0.6, 0.8 and 1.0 mm (paper 34, 42, 38). Every trial of the VGAT thalamic sessions has a 1.3 s delay, so W1 ends with the delay epoch and holds no time after the photostimulus (W1 on 1.3 s delays only gives the same values). In 100 ms bins the ratio is 21.6% in the bin from 1.4 s after pole onset, falls to 9 to 11% by the end of the delay and returns to 100% in the response epoch.
How the test differed from the paper’s method
- Distance is measured from the files' photostimulus site (AP -1.5, ML 0.9, DV -4.1 mm), the same nominal coordinate in every session, not from the fibre tip as placed in each mouse.
- The files give the photostimulus power as 2.23 mW; the paper gives 10 mW.
- Unit classes under readings P1 (computed width) and P2 (file label) of convention C.
ReproducedResults, Preparatory activity in thalamus, paragraph 1; Figure 2e
Approximately equal numbers of selective ALM neurons preferred contra and ipsi movements.
Approximately equal numbers of neurons preferred contra- or ipsi-lateral movements ( Fig. 2e )
Results, Preparatory activity in thalamus, paragraph 1; Figure 2e
Under all 32 readings the contra-preferring fraction of selective ALM neurons is 0.495 to 0.543, the binomial test does not reject one half, and the fraction lies within 0.45 to 0.55; both readings of the criterion hold everywhere.
How the test differed from the paper’s method
- Unit classes under readings P1 and P2 of convention C (answer of 2026-10-20). Outcome unchanged.
ReproducedResults, Thalamus drives ALM preparatory activity, paragraph 3; Figure 3e, f
Photoinhibition reduced thalALM activity 2.5 ms after photostimulus onset, and ALM spike rates were reduced 8.2 ms after onset, a latency difference consistent with direct thalamocortical drive.
Photoinhibition reduced thal ALM activity 2.5 ± 0.8 ms after photostimulus onset (mean ± SEM, activity reduction onset detected by 3 × std, Fig. 3f ). Hyperpolarization of ALM neurons and reduction of ALM spike rates were detected 7.9 ± 1.7 ms and 8.2 ± 1.5ms after photostimulus onset
Results, Thalamus drives ALM preparatory activity, paragraph 3; Figure 3e, f
Under reading O2 with correct and incorrect trials (T1), the thalamic onset is 2.0 ms (SEM 1.08 with P1, 0.96 with P2) and the ALM onset 9.0 ms (P1, SEM 3.68) or 7.0 ms (P2, SEM 3.16), within 2.5 +/- 0.8 and 8.2 +/- 1.5 ms, ALM later. Under O1 the ALM onset is 0 ms in three of four cases (the first 1 ms bin already lies 3 SD below control) and under T2 with O2 it is 6 ms; those readings fail. The pass rests on one reading of four, and the onset rule is fragile. On photostimulus PSTHs simulated from each unit's control PSTH with no effect it reports an onset below 10 ms in 30 to 48% of runs, and with a step at 8 ms it reports one before 8 ms in 18 to 43% and within 6.5 to 9.5 ms in 60 to 84%. The ALM onset of 0 ms under O1 is such a false onset.
How the test differed from the paper’s method
- The photostimulus onset is the delay onset of the trial table (1.3 s after pole onset), as the extracellular files hold no laser trace; the data check shows thalamic units fall within 5 ms of it.
- Populations are those of fig3f-thalamus-silenced (reading a) and fig3-alm-collapse, which the answers of 2026-10-20 leave unchanged; outcome unchanged.
ReproducedResults, Thalamus drives ALM preparatory activity, paragraph 4; Figure 4g-i
Photoinhibition of contralateral ALM barely changed the membrane potential of ALM neurons (1.4 mV).
photoinhibiting contralateral ALM, the anatomically strongest input to ALM, barely changed membrane potential (-1.4 ± 0.5 mV; mean ± SEM, n = 9 cells)
Results, Thalamus drives ALM preparatory activity, paragraph 4; Figure 4g-i
Under M the 9 cells change by -1.60 +/- 0.64 mV, within -1.4 +/- 0.5; under the laser onset (L) by -1.38 +/- 0.58 mV, within; under E by +0.86 mV, where the window precedes the change.
How the test differed from the paper’s method
- The photostimulus onset is placed by readings E and M, and by the laser onset (L, question 7).
ReproducedResults, Thalamus drives ALM preparatory activity, paragraph 4; Figure 4b vs 4e, h
thalALM drives ALM more strongly than M1 or contralateral ALM, the other reciprocally connected structures tested.
Altogether thal ALM drives ALM directly and more strongly than other reciprocally connected structures (cf. Fig. 4b vs Fig. 4e,h )
Results, Thalamus drives ALM preparatory activity, paragraph 4; Figure 4b vs 4e, h
Under M thalamus photoinhibition (-16.83 mV) is more negative than M1 (-2.75 mV) and contralateral ALM photoinhibition (-1.60 mV), p = 5.2e-8 and 7.9e-9 (Student) and 5.0e-8 and 3.4e-10 (Welch); under the laser onset (L), -13.68 against -3.26 and -1.38 mV, p = 1.3e-6 and 8.1e-8 (Student) and 6.7e-7 and 3.7e-9 (Welch). Readings (a) and (b) both hold under M and L. Under E, where the window precedes the change in all three groups, neither holds (p = 0.41 to 0.58).
How the test differed from the paper’s method
- The photostimulus onset is placed by readings E and M, and by the laser onset (L, question 7).
ReproducedResults, Preparatory activity in thalamus, paragraphs 1 and 2; Figure 2e, f; Extended Data Figure 4c, f
In ALM and in VM/VAL, population selectivity emerged in the sample epoch and increased throughout the delay epoch; it peaked at the beginning of the response epoch in ALM and during the response epoch in VM/VAL.
Selectivity for movements emerged in the sample epoch, increased throughout the delay epoch, and reached a maximum at the beginning of the response epoch ... Selectivity emerged in the sample epoch, increased throughout the delay epoch, and reached a maximum during the response epoch
Results, Preparatory activity in thalamus, paragraphs 1 and 2; Figure 2e, f; Extended Data Figure 4c, f
With VM/VAL P3 (258 units, t2) and ALM P2 (810 units, t2), both areas meet the criterion. Selectivity emerges in the sample epoch (first window above twice its SEM centred at 0.1 s in ALM and 0.3 s in VM/VAL), is higher at the end of the delay than at its start (ALM 0.85 to 1.10, VM/VAL -0.05 to 0.38 spikes/s), and peaks in the response epoch (3.0 and 3.3 s). Six combinations of readings meet the criterion in both areas, all with VM/VAL P3 (all neurons with t2, selective neurons with t1 and t2) and ALM P1 or P2. With VM/VAL P1 or P2, selectivity is lower at the end of the delay than at its start in every reading (P2, all neurons, t1, 0.83 against 0.17), and with P3, all neurons and t1 as well (0.57 against 0.53), so the VM/VAL increase holds in 3 of 12 VM/VAL readings. The ALM peak lies 0.4 to 0.6 s after the go cue, not within its first 200 ms. The answers of 2026-10-20 keep both trial readings (t1, t2) and add P3.
How the test differed from the paper’s method
- Preference from total spike counts over the three epochs (convention S) in a random half of the trials; windows of 200 ms stepped by 100 ms; only trials with a 1.3 s delay whose spikes are stored to 4.1 s after pole onset.
BlockedResults, Thalamus drives ALM preparatory activity, paragraph 1; Figure 3d· Methods underspecified
Thalamus photoinhibition reduced ALM activity across all cortical layers.
across all cortical layers ( Fig. 3d )
Results, Thalamus drives ALM preparatory activity, paragraph 1; Figure 3d
By the sign of the change (m1), inhibited neurons outnumber activated ones in every depth bin under every reading (12 of 12 bins of 100 um, 7 of 7 of 200 um), and the population rate under photoinhibition is at most 0.205 of control in any bin. Counting significant changes (m2), 1 to 3 bins fail with Student's test and 1 with Welch's, each holding 1 to 4 neurons with no significant change in either direction (0 against 0). The criterion blocks a claim whose readings disagree. The reduction is seen at every depth sampled (0.1 to 1.3 mm). The answers of 2026-10-20 leave its units (VGAT at 10 mW, P1 and P2) as they were; outcome unchanged.
How the test differed from the paper’s method
- Depth is the files' `unit_y`, described only as "y-coordinate of this unit (mm)"; whether it is depth below the pia is not stated.
BlockedResults, Thalamus drives ALM preparatory activity, paragraph 1; Figure 3b· Methods underspecified
ALM neurons with preparatory activity, peri-movement selectivity or both were equally inhibited by thalamus photoinhibition.
neurons showing preparatory activity, peri-movement selectivity or both were equally inhibited; Fig. 3b , P > 0.2, t -test
Results, Thalamus drives ALM preparatory activity, paragraph 1; Figure 3b
What the paper reports
- t-test, p > 0.2
Under the first selectivity reading both measures find a difference between groups, the rate difference (p < 0.0001, peri-movement only against both) and the ratio (p = 0.016 with P1, 0.010 with P2, T1); the claim would not be reproduced there. Over all 64 combinations of selectivity, trial and unit readings, the difference measure finds a pair at p < 0.05 in all and the ratio measure in 13, so the measures disagree in most readings, which the criterion calls methods underspecified. Mean ratios are about 0.13 (preparatory only), 0.04 to 0.05 (peri-movement only) and 0.02 (both). The answers of 2026-10-20 leave its units as they were; outcome unchanged.
How the test differed from the paper’s method
- Groups and populations under the selectivity readings of convention S and P1, P2.
BlockedResults, ALM drives thalamic preparatory activity, paragraph 1; Figure 6b· Methods underspecified
VM/VAL neurons selective during the sample, delay or response epochs were equally inhibited by ALM photoinhibition.
VM/VAL neurons selective during sample, delay or response epochs were equally inhibited ( Fig. 6b , P > 0.2, t -test)
Results, ALM drives thalamic preparatory activity, paragraph 1; Figure 6b
What the paper reports
- t-test, p > 0.2
With groups (a), a pair differs at p < 0.05 in 10 to 16 of the 16 selectivity readings for the rate difference and 7 to 9 for the ratio; with groups (b), in 1 to 10 readings for the difference and 0 or 1 for the ratio. The readings neither all find nor all miss a difference. The answer of 2026-10-20 adds P3 (all VM/VAL units, 201); under it groups (a) differ in 10 or 11 readings for the difference and 7 or 8 for the ratio, groups (b) in 1 to 5 and 0, so the readings still disagree. Outcome unchanged.
BlockedResults, ALM drives thalamic preparatory activity, paragraph 2; Extended Data Figure 8f, g· Metadata insufficient
Photoinhibiting vibrissal M1 did not reduce thalALM activity (Results), or reduced VM/VAL activity less than ALM photoinhibition in the same neurons (Extended Data Fig. 8g legend).
photoinhibiting the nearby vibrissal primary motor cortex (vM1), which projects to parts of VA/VL, did not cause a reduction in thal ALM activity ( Extended Data Fig. 8d-g )
Results, ALM drives thalamic preparatory activity, paragraph 2; Extended Data Figure 8f, g
The files name no vM1 site, and the answer of 2026-10-20 says the paper does not say which deposited condition is vM1, so the claim stays blocked. The answer fixes the comparison at ALM 1.5 mW, with matched power a variant. Taking photostimulation at galvo (0.5, 0.5) mm in mouse anm00264943 (6 sessions, 127 trials, 0.79 mW) as vM1 only to show the data, in 41 VM/VAL units (P2, T1) it lowers activity to 89.6% of control (paired p = 0.058; T2 96.8%, p = 0.62), and ALM photoinhibition at 1.5 mW lowers the same units to 43.3%, a larger reduction (p < 0.0001). (a) and (b) hold together under P2 with T1 and T2 and under P3 with T2; under P3 with T1 (48 units) the site reduces activity significantly (p = 0.040). In the variant with ALM at 0.79 mW the ALM reduction is no larger (p = 0.73). The paper's 46 cells come from 3 mice, these from 1.
BlockedMethods, Electrophysiology data analysis, paragraph 4; Extended Data Figure 4j, 10b· Data missing from the dataset
Putative GABAergic SNr neurons had high spike rates (40.9 spikes per s) compared with the other SNr units (23.4), and most of them (152/181) were selective for trial type, like ALM and VM/VAL.
These units have high spike rates (40.9 ± 21.5, mean ± SD, n = 181). For comparison, neurons with longer spike widths have lower spikes rates (23.4 ± 17.0, mean ± SD, n = 46)
Methods, Electrophysiology data analysis, paragraph 4; Extended Data Figure 4j, 10b
Spikes are stored only to about 4.8 s after pole onset in each 9 s trial and not between trials, so neither reading of the mean rate can be computed as defined; as computed, both give 20.7 (P1) or 22.0 (P2) spikes/s for the putative GABAergic units against 40.9, biased low because they divide by time without stored spikes. Over the stored part of each trial (a variant) the same units fire 36.0 (P1) or 38.1 (P2) spikes/s and the other units 13.1 (10 units, P1) or 20.3 (35 units, P2) against 23.4. Only 10 (P1) or 35 (P2, most labelled 'Cannot be determined') other SNr units are deposited against 46. The selectivity part is tested and within tolerance under P1, 139 of 171 under the first reading (137 to 143 over the c1 readings), against 152 +/- 15; under P2 114 to 122 of 146, below it.

Not tested in a notebook 12

9 data missing from the dataset, 1 metadata insufficient, 2 methods underspecified

BlockedResults, Thalamus is required for motor preparation, paragraph 3; Figure 1d· Data missing from the dataset
Unilateral photoinhibition of thalALM during the delay epoch lowered performance on contra trials and caused a small, non-significant improvement on ipsi trials.
Similar to ALM, unilateral photoinhibition of thal ALM during the delay epoch reduced performance for contralateral trials and caused a small improvement in performance for ipsilateral trials ( P < 0.001 for contralateral trials, not significant for ipsilateral trials, paired t -test
Results, Thalamus is required for motor preparation, paragraph 3; Figure 1d
What the paper reports
- contra trials, photoinhibition against control: paired t-test, p < 0.001 (n = 9 sessions, 4 mice)
The only Gad2 sessions (genotype 'Ai35D x GAD2_Cre', 9 thalamic recordings of 2 mice) carry no photostimulus trial, and there are no behaviour-only sessions with thalamus photoinhibition in Gad2-IRES-Cre mice.
BlockedResults, Thalamus is required for motor preparation, paragraph 3; Extended Data Figure 2c· Metadata insufficient
Photostimulating TRN axons in thalamus of Gad2-IRES-Cre mice reduced thalALM spike rates, recorded near VM/VAL, to 67.8% of control during the delay epoch.
Photostimulating axonal terminals of TRN GABAergic neurons in thalamus reduced spike rates in thal ALM (recorded near VM/VAL, to 67.8% during the delay epoch
Results, Thalamus is required for motor preparation, paragraph 3; Extended Data Figure 2c
The 9 Gad2 thalamic sessions hold 84 units, the 26 + 41 + 17 optrode units of Extended Data Fig. 2c, but every trial has photostimulus period 'N/A' and power 0, so photoinhibition trials cannot be told from control trials. The genotype is given as 'Ai35D x GAD2_Cre', where the paper used Gad2-IRES-Cre with ChR2 injected in TRN.
BlockedResults, Thalamus is required for motor preparation, paragraph 3; Extended Data Figure 3· Data missing from the dataset
Small doses of muscimol infused near VM/VAL produced an ipsilateral bias in licking responses, while infusions outside thalALM, anterior or dorsal (1.1 to 1.6 mm anterior or 1.6 to 1.9 mm dorsal to VM/VAL in the Results; about 0.3 to 0.8 mm anterior to VM/VAL and 0.2 to 0.5 mm dorsal to MD in the legends), did not, even at ten-fold higher doses.
Inhibition near VM/VAL produced ipsilateral bias in the licking response ( Extended Data Fig. 3a, b ). Inhibition outside of thal ALM (1.1 - 1.6 mm anterior or 1.6-1.9 dorsal to VM/VAL) did not cause ipsilateral bias, even at 10-fold higher doses
Results, Thalamus is required for motor preparation, paragraph 3; Extended Data Figure 3
What the paper reports
- performance change near VM/VAL against control: Student's t-test, p < 0.05 (n = 6 infusion sites)
The dandiset holds no muscimol sessions or infusion metadata.
BlockedResults, Thalamus drives ALM preparatory activity, paragraph 2; Extended Data Figure 5c-h· Data missing from the dataset
The hyperpolarization of ALM neurons during thalamus photoinhibition was caused mainly by a loss of excitation, not by an increase in inhibition; it was larger with negative current injection.
Manipulating the membrane potential during recording revealed that hyperpolarization was mainly caused by a reduction in excitation, not an increase in inhibition
Results, Thalamus drives ALM preparatory activity, paragraph 2; Extended Data Figure 5c-h
What the paper reports
- slope of the regression of the Vm change on baseline Vm: bootstrap, p < 10⁻⁴
The dandiset holds no non-behaving recordings and no current-injection experiments during TRN photostimulation in Gad2-IRES-Cre mice; all 36 cells are behaving, with no injected current around the photostimulus.
BlockedResults, Thalamus drives ALM preparatory activity, paragraph 3; Extended Data Figure 6e-g· Data missing from the dataset
The latency of ALM hyperpolarization is too short for the thalamus to M1 to ALM pathway; thalamus photoinhibition hyperpolarized M1 neurons after 11.0 ms and M1 photoinhibition hyperpolarized ALM neurons after 13.9 ms, so that pathway would give a latency difference above 14.3 ms against the 5.4 ms observed.
The thalamus → M1 → ALM pathway would produce a latency difference of > 14.3 ms, too slow by > 9 ms to explain the hyperpolarization in ALM
Results, Thalamus drives ALM preparatory activity, paragraph 3; Extended Data Figure 6e-g
The dandiset holds no whole-cell recordings of M1 neurons (every whole-cell recording is in ALM), so the M1 onset after thalamus photoinhibition cannot be measured. The other onset, ALM after M1 photoinhibition, can be measured from the laser trace the whole-cell files hold (reading L, question 7) and is 8.7 +/- 4.6 ms (V1) or 27.3 +/- 6.6 ms (V2), against 13.9 +/- 3.2; the claim needs both.
BlockedExtended Data Figure 6c· Data missing from the dataset
Photoactivating thalamus in Olig3-cre x Ai32 mice depolarized ALM neurons after 3.6 ms, an estimate of the thalamocortical conduction delay.
The time between photostimulation onset and depolarization onset was 3.6 ± 1.1 ms ( n = 9 cells)
Extended Data Figure 6c
No Olig3-cre x Ai32 recordings are in the dandiset (genotypes are PV-Cre x Ai32, VGAT-ChR2-EYFP, Ai35D x GAD2_Cre and C57BL6).
BlockedResults, Thalamus drives ALM preparatory activity, paragraph 3; Extended Data Figure 6d; Supplementary information· Methods underspecified
A model ALM neuron with 200 thalamic inputs predicts at least 1.8 ms between full loss of thalamic input and a detectable change in membrane potential.
Even when 100% of the input neurons were inhibited, we expect to observe latency of 1.8 ± 0.7 ms
Results, Thalamus drives ALM preparatory activity, paragraph 3; Extended Data Figure 6d; Supplementary information
A simulation whose parameters are in the Supplementary Information, which the claims file was not written from; no data in the dandiset bear on it.
BlockedResults, Thalamus contributes to selectivity in ALM, paragraph 1; Figure 5b· Data missing from the dataset
Weak thalamus photoinhibition (20-fold weaker, 0.5 mW) significantly changed the spike rate of 83 of 160 ALM neurons, with a mean reduction of 1.2 spikes per s; 39% of the changed neurons (32/83) increased their rate.
83 /160 of ALM neurons showed statistically significant changes in mean spike rate during photoinhibition ( t -test; average reduction of spike rate, 1.2 ± 0.2 spikes per s, mean ± SEM, n = 160; Fig. 5a, b ). A substantial fraction of these neurons (39%, 32/83) showed increases in spike rate
Results, Thalamus contributes to selectivity in ALM, paragraph 1; Figure 5b
No ALM session has 0.5 mW thalamus photoinhibition; the VGAT ALM sessions use 10 mW only and the PV-Cre x Ai32 ALM sessions 10, 22.5 and 45 mW. The dandiset holds 966 of the 1214 ALM units and 47 of 57 sessions, consistent with the Fig. 5 sessions (sorted with JRclust) being absent.
BlockedResults, Thalamus contributes to selectivity in ALM, paragraph 1; Figure 5c; Extended Data Figure 7a, b· Data missing from the dataset
Weak thalamus photoinhibition reduced the selectivity of ALM neurons, both contra- and ipsi-preferring; the change in selectivity was anti-correlated with selectivity (slope -0.33, r = -0.75), even in neurons without a change in mean rate.
This unequal spike rate change reduced the selectivity of ALM neurons in both contra- and ipsi- preferring neurons ( Fig. 5c, d and Extended Data Fig. 7a, b )
Results, Thalamus contributes to selectivity in ALM, paragraph 1; Figure 5c; Extended Data Figure 7a, b
Depends on fig5-weak-photoinhibition-rates, whose sessions are not in the dandiset.
BlockedResults, Thalamus contributes to selectivity in ALM, paragraph 2; Extended Data Figure 7c-e; Supplementary information· Methods underspecified
The data are inconsistent with linear network models with specific or non-specific thalamocortical connectivity, but do not exclude non-linear models of either kind.
Our data are inconsistent with linear models with either specific (like-to-like) or non-specific (all-to-all) thalamocortical connectivity. However, our data do not exclude non-linear models
Results, Thalamus contributes to selectivity in ALM, paragraph 2; Extended Data Figure 7c-e; Supplementary information
A modelling result specified in the Supplementary Information, which the claims file was not written from.
BlockedResults, Thalamus is required for motor preparation, paragraph 2; Extended Data Figure 1d, f· Data missing from the dataset
ALM is reciprocally connected with ipsilateral thalamic nuclei, most consistently the VM/VAL complex; a retrograde injection into ALM labelled 17375 thalamic neurons, against 38062 in contralateral ALM and 26599 in M1.
Within thal ALM , the VM/VAL complex was most extensively and consistently labeled in experiments involving injections of multiple types of anterograde and retrograde tracers into ALM
Results, Thalamus is required for motor preparation, paragraph 2; Extended Data Figure 1d, f
The dandiset holds no histology or cell counts.
BlockedResults, ALM drives thalamic preparatory activity, paragraph 2; Extended Data Figure 8h, i· Data missing from the dataset
GABAergic neurons in ALM do not project to thalALM, so ALM photoinhibition does not inhibit thalamus directly.
Second, GABAergic neurons in ALM do not project to thal ALM ( Extended Data Fig. 8h, i ), excluding direct photoinhibition of thal ALM
Results, ALM drives thalamic preparatory activity, paragraph 2; Extended Data Figure 8h, i
The dandiset holds no histology.

Findings about the data

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

Data missingThe dandiset holds 966 of the paper's 1214 ALM units (47 of 57 sessions), 789 of 790 thalamic units (72 of 73 sessions) and 181 of 227 SNr units (18 of 23 sessions; apparently only the putative GABAergic units). Behaviour-only sessions, the muscimol experiments, the weak (0.5 mW) thalamus photoinhibition of Fig. 5, non-behaving whole-cell recordings with current injection, M1 whole-cell recordings, Olig3-cre recordings and the histology are not deposited.
Data missingSpikes are stored only from each trial's start to about 4.8 s after pole onset, although trials last about 9 s, and not between trials. The end is fixed to pole onset, not to the go cue, so in trials with delays above about 2 s the response epoch is not stored; spike rates over whole trials or sessions cannot be computed.
Description missingMost units hold spikes over only part of their session (median 70% of its trials; 1358 of 1936 units below 90%), and the Units table gives no observation interval, so the trials in which a unit was recorded must be inferred from its first and last spikes.
File format errorThe mean waveforms are cut short (29 to 38 samples at 25 kHz, trough at sample 15 to 19), so trough-to-peak widths above about 0.5 ms cannot be measured; in 78% of ALM units labelled 'Wide width spiking' the peak is the last sample.
Selection not documentedThe Units `cell_type` labels ('Wide width spiking', 'Narrow width spiking', 'Cannot be determined') have no stated rule; they cannot be derived from the deposited waveforms but give class sizes close to the paper's.
Description missingThe whole-cell trial tables leave every photostimulus column empty, and `stim_present` marks the photostimulus trials without description. The laser time is held only in an OptogeneticSeries (stimulus/presentation/photostim_on_*) whose description is 'no description'. It starts between the two delay onsets the trial times allow, 0.10 to 0.17 s after 1.3 s past pole_in_time and 0.03 to 0.10 s before 0.1 s past pole_out_time (1.40 s after pole onset in mouse anm300476, 1.47 s in the others), and it is on in one trial not marked by stim_present (cell_21_1).
Metadata errorIn the whole-cell file of cell_21_1 (anm300476) the trial start and stop times are shifted by one row against the pole and cue times (85 rows), and the table is not in time order.
Metadata errorThe whole-cell Vm series state unit 'mV' with conversion 0.001, but the values are already in mV; applying the conversion gives values a thousand times too small.
Metadata errorEvery electrode group, in thalamus and SNr as well as ALM, carries the ALM coordinates (AP 2.5, ML -1.5, DV 0); the SNr units of anm00261336 are placed at AP -1.56 mm, outside SNr; photostimulus sites give ML negative for the left hemisphere while units give it positive.
Description missingThe Units column unit_y is described only as "y-coordinate of this unit (mm)", with no reference, so whether it is depth below the pia is not stated; session_start_time is the date of conversion (all sessions on 2017-06-27), not of recording, and no sorter is named, so the ALM sessions missing from the dandiset cannot be identified.
Description missingThalamic units have no nucleus label (brain_subregion 'N/A'), so VM/VAL units can be selected only by distance from a centre the paper does not give in coordinates.
Metadata errorThe 9 Gad2 thalamic sessions give the genotype 'Ai35D x GAD2_Cre' (an Arch line), where the paper used Gad2-IRES-Cre with ChR2 injected in TRN, and mark no photostimulus trial, though their 84 units match the optrode units of Extended Data Fig. 2c.
Metadata errorThe VGAT-ChR2-EYFP thalamic sessions give the photostimulus power as 2.23 mW where the paper gives 10 mW (Extended Data Fig. 2d); the thalamic fibre site is one nominal coordinate in every session.
Description missingPhotostimulation at galvo position (0.5, 0.5) mm in mouse anm00264943, interleaved with ALM photoinhibition, is not named; trials labelled 'delay' with power 0 are not explained; the trial columns photo_stim_period, photo_stim_type and is_good have no description.
OtherThe dandiset holds data the paper does not describe, 21 ALM sessions of PV-Cre x Ai32 mice with thalamic photostimulation at 10, 22.5 and 45 mW and 12 thalamic sessions of C57BL6 mice without photostimulation.

Notebooks

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

Reproduction Persistent activity in a frontal thalamocortical loop (Guo et al. 2017)

Tests the findings of Guo, Inagaki, Daie, Druckmann, Gerfen and Svoboda (2017) on dandiset 000009, which holds silicon-probe recordings from ALM, thalamus and SNr during a delayed licking task with optogenetic photoinhibition, and whole-cell recordings of ALM neurons during thalamus, M1 or contralateral ALM photoinhibition. It tests behaviour under ALM photoinhibition, unit classes, trial-type selectivity, the collapse of ALM and thalamic activity under photoinhibition of the other area, onset latencies, SNr responses and membrane potential measures. Behaviour-only, muscimol, weak photoinhibition, non-behaving whole-cell and anatomical data are not in the dandiset. All files

Reproducing Guo et al. (2017) from dandiset 000009
reproduce_guo_2017.ipynbExecuted, with outputs · Source

Verification

Round 1
2026-10-20
Agreed with 29 of 39 outcomes. Independent agent run by CatalystNeuro (same harness, separate context, did not write the notebook). Reviewed: Round 1. The executed notebook of the author-answers run with the conclusion cells and the summary removed, the helper module and the claims file, with the paper. The verifier queried the archive metadata and tested helper functions on synthetic inputs. Its report had no attempt record and gave its own outcome for each claim; 29 of 39 agree with the record as it then stood, 10 do not (it would give not reproduced where the record gave partially reproduced).
- task-performance: the verifier’s outcome was “Not reproduced”. No reading has all three values within tolerance. The closest, whole-cell trials pooled, gives 77.4% correct and 9.65% lick-early, but 1.55% no-response against 1.1 +/- 0.11.
- alm-unit-classes: the verifier’s outcome was “Not reproduced”. 966 units against 1214 +/- 121, and pyramidal and fast-spiking counts are out under P1 and P2. The P2 fractions do match (0.839 against 0.829; 0.125 against 0.137); about 250 units are simply not deposited.
- fig2e-alm-selectivity: the verifier’s outcome was “Not reproduced”. 0 of 32 readings meet the criterion on absolute counts. The fraction matches (P2 0.70), and scaled to 1006 units control/R2/c1 gives 702/156/268/277, within tolerance. The shortfall is missing units. The response windows reach beyond where the stored spikes end, about 5 s after pole onset, and whether storage ends relative to pole onset or to the cue was never shown; this bears on every selectivity claim (problem 2).
- thalamus-unit-classes: the verifier’s outcome was “Not reproduced”. 789 units and 682 putative thalamic (P2) are within tolerance. Within 0.4 mm gives 259 (allowed 282 to 344) and VM/VAL 222 (allowed 265 to 325). Reading L1 has no label.
- fig2f-vmval-selectivity: the verifier’s outcome was “Not reproduced”. Counts are out under all 64 readings. Under P2 and P3 the selectivity types differ from ALM (2 x 4 test p = 0.0003 to 0.04). Only P1 (n = 20) passes the chi-squared tests. The response windows may reach beyond the stored spikes, which bears on every selectivity claim (problem 2).
- fig3f-thalamus-silenced: the verifier’s outcome was “Not reproduced”. Q1 is 0.9 to 1.1% against 2 +/- 0.2, and n is 89 to 113 against 148 +/- 15. Reading (b), VM/VAL, has 0 units.
- wc-thalamus-hyperpolarization: the verifier’s outcome was “Not reproduced”. Under E the change is +0.31 mV. Under M it is -16.83 +/- 1.24 mV, outside -18.7 +/- 1.1; the onset is not judged. Provisional, because the archive metadata list an OptogeneticSeries in all 36 whole-cell assets that the loader never reads (problem 1).
- wc-m1-small-effect: the verifier’s outcome was “Not reproduced”. Under M, Vm is -2.75 (within) but the rate change is +0.76 spikes/s against -1.4 +/- 0.6. Under E both values are out.
- fig6d-localized: the verifier’s outcome was “Not reproduced”. Group counts are out in all 12 readings, e.g. 111 against 160 and 26 against 46. Near < far does hold under P2 and P3.
- snr-photoinhibition: the verifier’s outcome was “Not reproduced”. The reduction is 2.1 to 2.4 spikes/s against 3.6 +/- 0.36 (5.6 to 6.3% against 8.3 +/- 0.83), and 181 SNr units are deposited against 227 +/- 23. The bootstrap p < 0.01 holds in most readings. GABA P2 uses every unit not labelled wide (168), where the claims file's P2 is "Narrow width spiking" (146) (problem 6).
Round 2
2026-10-21
Agreed with 30 of 39 outcomes. Independent agent run by CatalystNeuro (same harness, separate context, did not write the notebook). Reviewed: Round 2. The full executed notebook of the author-revise-1 run, with its conclusions and summary, the helper module, the attempt record, the README and the claims file, with the paper. The verifier checked the archive metadata of the 137 extracellular assets. It agrees with 30 of 39 outcomes and would give not reproduced for 9 given partially reproduced. No outcome changes in this round; for those 9 a change to not reproduced is proposed, for a third round.
- alm-unit-classes: the verifier’s outcome was “Not reproduced”. None of these criteria has a partial clause. Each says "Reproduced when ..." and nothing else, and no reading meets it (cell 25).
- fig2e-alm-selectivity: the verifier’s outcome was “Not reproduced”. None of these criteria has a partial clause. Each says "Reproduced when ..." and nothing else, and no reading meets it (cell 28). Problem 2 (spike storage) appears in no round-1 reason, only in the responses to this claim and fig2f-vmval-selectivity.
- thalamus-unit-classes: the verifier’s outcome was “Not reproduced”. None of these criteria has a partial clause. Each says "Reproduced when ..." and nothing else, and no reading meets it (cell 34).
- fig2f-vmval-selectivity: the verifier’s outcome was “Not reproduced”. None of these criteria has a partial clause. Each says "Reproduced when ..." and nothing else, and no reading meets it (cell 37).
- fig3f-thalamus-silenced: the verifier’s outcome was “Not reproduced”. None of these criteria has a partial clause. Each says "Reproduced when ..." and nothing else, and no reading meets it (cell 49).
- wc-thalamus-hyperpolarization: the verifier’s outcome was “Not reproduced”. No partial clause, as the record itself says. Firmly not reproduced; under the claims file's readings E gives +0.31 mV and M -16.83 mV, outside -18.7 +/- 1.1. The 73% figure that supports partial comes from reading L, which is not in the claims file (question 7 is open); even under L the Vm change (-13.68 mV) is out of tolerance. Also, M is 28 ms after the laser onset, not 27 (cell 14 gives the onset 0.0716 to 0.0720 s after pole_out_time); cell 92 says the onset latency is judged under L only while the record says L is shown pending question 7, and the claims file says the 7.9 ms onset cannot be judged under any reading it lists; the measured value lists 9.05 ms, the onset under M, which the notes call not judged.
- wc-m1-small-effect: the verifier’s outcome was “Not reproduced”. No partial clause, as the record itself says. The spike-rate change is out of tolerance under every reading (+0.94, +0.76 and -0.01 spikes/s against -1.4 +/- 0.6).
- fig6d-localized: the verifier’s outcome was “Not reproduced”. None of these criteria has a partial clause. Each says "Reproduced when ..." and nothing else, and no reading meets it (cell 67). Cell 68 and the notes say the group counts (240 to 286) match the paper's 250 near the centre, but 286 (u2, T1) is outside 250 +/- 25.
- snr-photoinhibition: the verifier’s outcome was “Not reproduced”. None of these criteria has a partial clause. Each says "Reproduced when ..." and nothing else, and no reading meets it (cell 73). The README states as fact "its 181 putative GABAergic SNr units", while cell 17 counts 171 under P1 and 146 under P2 out of 181; the record hedges this with "apparently".
Round 3
2026-10-21Agreed with 9 of 9 outcomes. Independent agent run by CatalystNeuro (same harness, separate context, did not write the notebook). Reviewed: Round 3. The 9 proposed changes of outcome (each partially reproduced to not reproduced) and the analysis behind them in the executed notebook after the round-2 revision (cells 25, 28, 34, 37, 49, 67, 73, 88 and 93), with the claims file and the paper. Nothing else was reviewed. Report in transcripts/000009/verifier-3-report.md.

Method differences

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

The photostimulus onset is placed by readings E and M, and by the laser onset (L, question 7).wc-m1-small-effect wc-contra-alm-small-effect thalamus-stronger-than-cortex
The dandiset holds recording sessions only; the paper's averages may come from behaviour sessions that are not deposited.task-performance
Rows with no outcome and no pole times (5369 extracellular rows, type 'non-performing') are not counted as trials.task-performance
Tested on the behavioural trials of the 66 recording sessions (ALM, thalamic and SNr recordings), not on the paper's 84 behaviour sessions, which are not in the dandiset.fig1c-alm-photoinhibition-bias
Photoinhibition trials at 1.5 mW (the Methods' power, convention P); trials at 0.75 or 0.79 mW are pooled in a variant only.fig1c-alm-photoinhibition-bias
Distance is measured from the files' photostimulus site (AP -1.5, ML 0.9, DV -4.1 mm), the same nominal coordinate in every session, not from the fibre tip as placed in each mouse.vgat-thalamus-reduction
The files give the photostimulus power as 2.23 mW; the paper gives 10 mW.vgat-thalamus-reduction
Unit classes under readings P1 (computed width) and P2 (file label) of convention C.vgat-thalamus-reduction
The deposited mean waveforms are too short to measure the widths the Methods use; both the computed widths (P1) and the files' labels (P2) are shown, as convention C says.alm-unit-classes
Each unit's trials are those from the trial holding its first spike to the trial holding its last, as the files give no observation interval per unit; this applies to every unit analysis.fig2e-alm-selectivity
A window that ends after the last stored spike of a trial (spikes are stored to about 4.8 s after pole onset) is counted as missing for that trial; this applies to every unit analysis and touches 221 of 8991 kept ALM trials, 173 of 14997 thalamic and 19 of 3573 SNr for R1.fig2e-alm-selectivity
Unit classes under readings P1 and P2 of convention C (answer of 2026-10-20; P3 does not apply to ALM). Outcome unchanged.fig2e-alm-selectivity
Unit classes under readings P1 and P2 of convention C (answer of 2026-10-20). Outcome unchanged.fig2e-alm-contra-ipsi-balance
The files have no VM/VAL label (reading L1 cannot be applied); reading L2 uses the claims file's proxy centre.thalamus-unit-classes
Unit classes under readings P1 and P2 of convention C; the criterion excludes P3 for these counts (answer of 2026-10-20). Outcome unchanged.thalamus-unit-classes
VM/VAL by reading L2 only; unit classes under P1, P2 and P3 of convention C.fig2f-vmval-selectivity
Only the VGAT-ChR2-EYFP sessions at 10 mW, as the criterion says; the PV-Cre x Ai32 ALM sessions with thalamic photostimulation are shown as a variant, not judged.fig3-alm-collapse
Unit classes under P1 and P2 of convention C.fig3-alm-collapse
Depth is the files' `unit_y`, described only as "y-coordinate of this unit (mm)"; whether it is depth below the pia is not stated.fig3d-all-layers
Groups and populations under the selectivity readings of convention S and P1, P2.fig3b-selectivity-types-equally-inhibited
The files give the photostimulus power as 2.23 mW; one session was photostimulated in the sample epoch and drops out.fig3f-thalamus-silenced
Unit classes under P1 and P2 of convention C (Fig. 3, so P3 does not apply; all units are shown as a variant). Outcome unchanged.fig3f-thalamus-silenced
The photostimulus onset is the delay onset of the trial table (1.3 s after pole onset), as the extracellular files hold no laser trace; the data check shows thalamic units fall within 5 ms of it.fig3-onset-latencies
Populations are those of fig3f-thalamus-silenced (reading a) and fig3-alm-collapse, which the answers of 2026-10-20 leave unchanged; outcome unchanged.fig3-onset-latencies
The threshold is the Vm at the kink of convention V, averaged over every spike of the recording.wc-near-threshold
Spike peaks are located from the deposited spike-removed trace; the kink and clipping are computed by the claims file's rule and match the deposited trace.wc-near-threshold
The photostimulus onset is placed by readings E and M of the criterion, and by the onset of the deposited laser trace (L), a reading not in the claims file, shown pending question 7.wc-thalamus-hyperpolarization
The criterion has no partial clause; the hyperpolarization and the silencing are present under M and L at 73 to 90% of the paper's value, but no listed reading meets the criterion.wc-thalamus-hyperpolarization
The criterion has no partial clause; one of its two values is within tolerance under M and L.wc-m1-small-effect
VM/VAL by reading L2; ALM photoinhibition at 1.5 mW (convention P), other powers as a variant; unit classes P1, P2 and P3.fig6c-vmval-reduced
Unit classes under P1, P2 and P3 of convention C.fig6c-unmodulated-less-selective
Distances from the L2 proxy centre in the coronal plane; unit classes under P1 and P2, and all units (u2), which is reading P3 of the answer of 2026-10-20; under u2 activity within 1.0 mm falls to 56.4 to 58.3% (Q1) or 44.2 to 48.2% (Q3) in the 20 to 120 ms window. Outcome unchanged.fig6d-localized
SNr populations from snr-photoinhibition under P2, 5 trials, T1, Student.fig6e-onset-latencies
The answer of 2026-10-20 adds P3 for VM/VAL; its 201 units give onsets of 14 ms (T1) and 7 or 14 ms (T2), and no reading with them puts all three onsets within tolerance with SNr latest. Outcome unchanged.fig6e-onset-latencies
The 46 wider SNr units of the paper are not deposited; the dandiset's 181 units appear to be the putative GABAergic ones, shown also as a set of their own.snr-photoinhibition
Preference from total spike counts over the three epochs (convention S) in a random half of the trials; windows of 200 ms stepped by 100 ms; only trials with a 1.3 s delay whose spikes are stored to 4.1 s after pole onset.selectivity-time-course

Provenance

Attempt record: 000009-guo-2017-01
Paper: Maintenance of persistent activity in a frontal thalamocortical loop
Date: 2026-10-20
Made by: An agent, run by CatalystNeuro
Dataset: DANDI:000009
Version: 0.220126.1903 (doi:10.48324/dandi.000009/0.220126.1903)
Run on: GitHub Actions
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.

← DANDI:000008 DANDI:000010 →