Claims 48
From the paper A brainstem map of orofacial rhythms · claims file · checked 2026-10-21: ready
- code-dataset-link
The dandiset is the data the paper's analysis code reads, so that the code's choices fixed in this file apply to it.
How it will be judged
Reproduced when process_mat.py's inputs can be built from the NWB files (spike times per trial, unit quality metrics, CCF coordinates, airflow at 25 kHz with its timestamps, side and bottom DeepLabCut jaw and tongue traces and likelihoods, 3D jaw and tongue tracks) and the count of units passing get_good_units over all files is within 10% of 20,300 (with or without the drift metric, V1). Partially reproduced when the inputs exist but the count is outside 10%.
- fig1d-lick-breath-frequencies
Mice licked at 6.23 +- 2 Hz (95% range 3.59 to 8.91 Hz) and breathed at lower frequencies, 4.02 +- 1.24 Hz (95% range 2.12 to 7.17 Hz).
Licking occurred at 6.23 ± 2 Hz (mean ± std, 95% range: 3.59 - 8.91 Hz) while breathing occurred at lower frequencies (4.02 ± 1.24 Hz, 95% range: 2.12 - 7.17 Hz; Figure 1d)
Results, Coordination of breathing, licking, and swallowing, paragraph 2; Figure 1dHow it will be judged
Neither the Methods nor the code say how the frequencies are taken. Readings (a) the inverse of every interval between consecutive tongue onsets, and between consecutive inspiration onsets, within trials, pooled over all sessions; (b) the same within lick bouts only (code bouts). Onsets as in the header. The 95% range is the 2.5th to 97.5th percentile. Reproduced when, under one reading, all six values are within tolerance; partially reproduced when the two means are and a bound is not.
- fig1f-inspiration-between-licks
Inspiration always occurred between licks and never coincided with licking; licking preceded inspiration at a fixed latency.
Inspiration always occurred between licks and thus never coincided with licking. Licking preceded inspiration at a fixed latency
Results, Coordination of breathing, licking, and swallowing, paragraph 2; Figure 1fHow it will be judged
The paper shows this in raster plots and gives no measure. Measure, from the paper's definitions, the fraction of inspiration onsets inside lick bouts that fall within a tongue protrusion (tongue onset to tongue offset, code get_movement_timing), per session. "Never" is read (a) literally, no such inspiration in any session, and (b) as a rounded statement, under 5% of inspirations in pooled data (a standard choice, named here). Reproduced when (b) holds; the result under (a) is reported. The latency from the last lick to the next inspiration is reported per number of licks in the cycle.
- fig1g-lick-delayed-before-slow-breath
For breath cycles with two licks, the second lick was delayed preceding a slower breath compared to a fast breath, in all mice (one-sample t-test, two-tailed, p < 0.001), so breathing influences the licking rhythm.
In all mice, the lick timing was significantly delayed preceding a slower breath compared to a fast breath (Figure 1g, p<0.001, one sample t-test, two-tailed)
Results, Coordination of breathing, licking, and swallowing, paragraph 2; Figure 1gHow it will be judged
The per-session peak difference of the header (Lick-to-breath timing, code get_peak_diff_lick_behv), slow minus fast. One-sample two-tailed t-test against 0 under O4 (a) sessions and (b) mice. Reproduced when the mean is positive with p < 0.001 under a reading and every mouse's mean difference is positive; partially reproduced when p < 0.001 with a positive mean but one or more mice are not positive.
- fig1f-licking-mode-switch
During slow breathing, licking frequency increased abruptly to add licks within the breathing cycle (1, 2, 3 or 4 licks between breaths), keeping the fixed lick latency before the next inspiration, so licking shifted between modes of oscillation with breathing frequency.
During slow breathing, when the next lick could not be delayed further to accommodate the next breath, licking frequency abruptly increased to add additional licks within the breathing cycle to maintain the fixed lick latency preceding the next inspiration
Results, Coordination of breathing, licking, and swallowing, paragraph 3; Figure 1fHow it will be judged
The paper shows this in a raster grouped by the number of licks in the cycle (code volcano_plot_licking). With the breath cycles of the header, pooled per session, reproduced when the median inter-breath interval rises strictly from 1 to 2 to 3 licks per cycle (and to 4 where a session has at least 20 such cycles) in most sessions (more than half), and the median latency from the last lick to the next inspiration differs by less than 20% between the groups (a standard choice, named here).
- fig1i-breath-delayed-before-slow-lick
Inspirations following licking were significantly delayed for slower licking frequencies, anticipating the next lick (one-sample t-test, two-tailed, p < 0.001), so licking influences breathing.
inspirations following licking were significantly delayed for slower licking frequencies to anticipate the next lick (Figure 1i, p<0.001, one sample t-test, two-tailed)
Results, Coordination of breathing, licking, and swallowing, paragraph 4; Figure 1iHow it will be judged
The per-session peak difference of the header (Breath-to-lick timing, code get_peak_diff_breath_behv), slow minus fast; one-sample two-tailed t-test against 0 under O4 (a) and (b). Reproduced when the mean is positive with p < 0.001 under a reading.
- ed1-unidirectional-models-fail
Neither a free-running breathing oscillator that sets lick timing nor a free-running licking oscillator that sets breath timing reproduces the observed lick-breath relationship, suggesting that the two oscillators are bidirectionally coupled.
This pattern of lick-breath temporal relationship also cannot be produced by a free-running licking oscillator unidirectionally setting the timing of breathing rhythm
Results, Coordination of breathing, licking, and swallowing, paragraph 4; Extended Data Figure 1d-i; Methods, Simulation of licking-breathing couplingHow it will be judged
Simulate as the Methods say, drawing intervals from the empirical frequency distributions of fig1d-lick-breath-frequencies, Gaussian timing noise SD 10 ms. The paper gives no value for the fixed latency or the number of simulated cycles; standard choices, named here, are the median latency from the last lick to the next inspiration (breathing model) and from the last inspiration to the next lick (licking model) in the data, and as many simulated cycles as the pooled data. Apply the two peak-difference measures of the header to each simulation. Reproduced when the breathing-driven model gives a Fig 1i difference whose mean is not positive or not significant (two-tailed t-test over simulated sessions of the data's sizes, p >= 0.05, a standard choice named here since the paper reports no test) and the licking-driven model fails likewise on the Fig 1g measure, while the data give both differences positive.
- ed2-rats-similar
Licking and breathing in freely moving rats showed a similar lick-breath temporal relationship.
Independent analysis of breathing and licking in freely moving rats yielded similar temporal relationship (Extended Data Figure 2)
Results, Coordination of breathing, licking, and swallowing, paragraph 4; Extended Data Figure 2How it will be judged
Rat data (7 rats, lickometer contacts, thermocouple inspiration onsets at 10% of onset amplitude after 1-20 Hz filtering, Methods). Licks from bouts shorter than 1.75 s left out as the Methods say. The two peak-difference measures of the header computed per rat with tongue contacts as licks. Reproduced when both differences are positive in a majority of rats and positive on average (one-sample t-test over rats, two-tailed, p < 0.05, a standard choice named here since the paper reports no test).
- ed1bc-licks-pause-at-swallows
Licking was paused during detected swallows but not during ordinary expirations, so the detected events are genuine swallows.
Licking was paused during swallowing. Bottom, lick timing aligned to all expirations as a control. Lick was not paused during expiration.
Extended Data Figure 1b-c; Results, Coordination of breathing, licking, and swallowing, paragraph 1How it will be judged
Swallows detected as in the header. Per session, the probability of a tongue onset in 3.4 ms bins from -0.3 to 0.3 s around swallow onsets and around all expiration onsets inside lick bouts. The paper does not define expiration onset; it is taken as the Hilbert phase crossing pi (a standard choice, named here), and the bins and window are plotting choices. Reproduced when the mean lick probability from 0 to 100 ms after onset is lower around swallows than around expirations in most sessions with at least 20 swallows, by a paired Wilcoxon signed-rank test, p < 0.05 (a standard choice, named here).
- fig1k-rhythms-follow-swallow-at-fixed-latency
Inspirations and licks followed swallowing at relatively fixed latencies despite different breath and lick timing before the swallow, suggesting that swallowing interrupted the ongoing rhythms.
Inspirations and licks followed swallowing at relatively fixed latencies despite different breath and lick timing before the swallow (Figure 1k and m)
Results, Coordination of breathing, licking, and swallowing, paragraph 5; Figure 1k and 1mHow it will be judged
Swallow events and ranking as in O5 (a) and (b). For licks and for breaths separately, reproduced when the difference in median latency from swallow onset to the first following lick (inspiration) between the top and bottom thirds is smaller than the difference in their median cycle -1 (as Figure 1k right compares the two thirds) under a reading.
- fig1l-lick-interval-preserved
The inter-lick interval before and after swallowing was preserved (one-sample t-test, two-tailed, P = 0.649).
The inter-lick-interval before and after swallowing was preserved (Figure 1l, P=0.649, one-sampled t-test, two-tailed)
Results, Coordination of breathing, licking, and swallowing, paragraph 5; Figure 1lHow it will be judged
A claim of absence, judged on the paper's test. Delta(1,-1) per swallow event as in the header, bottom third by O5 (a), the code's selection, pooled across sessions. Reproduced when the one-sample two-tailed t-test against 0 gives p >= 0.05; not reproduced when p < 0.05. The mean Delta(1,-1) is reported, and O5 (b), (c) and (d) are shown as variants.
- fig1l-lick-cycle-prolonged
The lick cycle in which swallowing occurred was significantly prolonged, by 33.3 ms (one-sample t-test, two-tailed, P < 0.001).
whereas the lick cycle in which swallowing occurred was significantly prolonged by 33.3 ms (Figure 1l, P<0.001, one-sampled t-test, two-tailed)
Results, Coordination of breathing, licking, and swallowing, paragraph 5; Figure 1lHow it will be judged
Delta(0,-1) per swallow event, readings O5 (a) to (d). Reproduced when, under one reading, the mean is within tolerance and the one-sample two-tailed t-test gives p < 0.001; partially reproduced when the test holds and the mean is outside tolerance under every reading. Under O5 (a) the selection of the longest cycles 0 favours a positive Delta(0,-1); the result under (b) and (c) is reported beside it.
- fig1n-breath-interval-preserved
The inter-breath interval before and after swallowing was preserved (P = 0.225).
P=0.225 for inter-breath-interval before vs. after swallowing
Results, Coordination of breathing, licking, and swallowing, paragraph 5; Figure 1nHow it will be judged
A claim of absence, judged on the paper's test, as fig1l-lick-interval-preserved with inspiration onsets and the breath limits of the header (cycle 0 between 1/15 and 1/2 s). Reproduced when p >= 0.05 under O5 (a); not reproduced when p < 0.05. The other readings are variants.
- fig1n-breath-cycle-prolonged
The breath cycle in which swallowing occurred was significantly prolonged, by 48.0 ms (one-sample t-test, two-tailed, P < 0.001).
P<0.001 for inter-breath-interval in which swallowing occurred, one-sample t-test, two-tailed, prolonged by 48.0 ms
Results, Coordination of breathing, licking, and swallowing, paragraph 5; Figure 1nHow it will be judged
As fig1l-lick-cycle-prolonged, with breaths.
- units-count
Across 339 insertions in 31 mice, 20,300 single units passed quality control (60 +- 22 per probe), 5,700 in the cerebellum, 3,700 in the pons and 10,900 in the medulla.
we obtained 20,300 clear single units, including 5,700 units in the cerebellum, 3,700 units in the pons, and 10,900 units in the medulla
Results, Brainstem activity maps of orofacial rhythms, paragraph 2; Methods, Preprocessing, spike sorting, and quality metrics, paragraph 2How it will be judged
Units by the quality filter of the header, V1 both ways; region from the unit's CCF label (cerebellum CB, pons P, medulla MY, with their subregions). Insertions and mice are counted exactly. Extended Data Figure 4c gives 19,000 units with amplitude > 150 uV, a second reading of the total; a count within tolerance of either is reported as such.
- region-spike-rates
Neurons in the cerebellum (22 +- 29 spikes/s) and vestibular nucleus (38 +- 32) fired faster than neurons in the medulla (11 +- 14).
Neurons in the cerebellum and the vestibular nucleus exhibited higher spike rates than neurons in the medulla (22 ± 29 spikes/s, mean ± std, for cerebellum, 38 ± 32 spikes/s for vestibular nucleus, and 11 ± 14 spikes/s for medulla
Methods, Preprocessing, spike sorting, and quality metrics, paragraph 2; Figure 2eHow it will be judged
The rate is not defined; readings (a) the quality metric avg_firing_rate; (b) spikes over the session's trial time. Vestibular nucleus is the CCF VNC; medulla is MY with (i) and (ii) without VNC. Reproduced when the three means are within tolerance under one reading.
- fig2-licking-breathing-tuned-counts
Units showed significant phase tuning to licking (3551 of 20,300) or breathing (1015 of 20,300), but rarely to both (182 of 20,300).
Individual units exhibited significant phase tuning to either licking (3551/20300), or breathing (1015/20300), but rarely to both (182/20300)
Results, Brainstem activity maps of orofacial rhythms, paragraph 3How it will be judged
Phase tuning and MI as in the header, tuned at P < 0.001 and MI > 0.6, readings V2 (a) and (b) and V3 (a) and (b). The text does not say whether the licking and breathing counts include the 182 units tuned to both; readings (i) they include them; (ii) they exclude them. Reproduced when the three counts are within tolerance under (i) or (ii), for every reading of V2 and V3; partially reproduced when the counts are not but both-tuned units are under 10% of licking-tuned units and under 25% of breathing-tuned units, for every reading of V2 and V3 (a standard choice, named here; the paper's ratios are 5% and 18%, and the Methods say only that the relative proportions were robust to the MI criterion). The code's choices (order 5, 100 resamples) are shown as variants.
- fig2-swallow-modulated-counts
Of 18,485 units, 427 were modulated by swallowing only, 175 by both licking and swallowing and 55 by both breathing and swallowing.
Most units were modulated by swallowing only (427/18485), with a smaller group modulated by both licking and swallowing (175/18485) (Extended Data Figure 5f). Another small subset of units with swallowing-related activity were also modulated by breathing (55/18485)
Results, Brainstem activity maps of orofacial rhythms, paragraph 3; Extended Data Figure 5e-fHow it will be judged
Swallow modulation as in the header, readings V4 (a) and (b). The denominator 18,485 is not defined; readings (a) units of sessions with at least one detected swallow; (b) units with at least 20 swallows (code). "Swallowing only" is read (i) not licking- or breathing-tuned; (ii) not licking-tuned. Reproduced when the four numbers are within tolerance under one combination of (a) or (b) with (i) or (ii), for both readings of V4.
- fig2f-breathing-units-location
Units encoding breathing clustered near the preBotzinger complex and in the posterior ventral IRN.
Units encoding breathing were near the preBötzinger complex 22, as well as posterior ventral IRN
Results, Brainstem activity maps of orofacial rhythms, paragraph 4; Figure 2fHow it will be judged
The paper shows a map and gives no test. Regions by O3 (a) and (b). Reproduced when the fraction of breathing-tuned units among the units of each of the two regions exceeds the fraction among all pons and medulla units, each by a one-sided Fisher exact test at p < 0.05 (a standard choice, named here), under both readings of O3.
- fig2g-licking-units-location
Licking-related units were broadly distributed, including the hypoglossal (12N) and trigeminal (5N) motor nuclei, extensive portions of the IRN and parts of the pons.
The regions containing licking-related activity were remarkably broad, including the tongue- and jaw-related motor nuclei (hypoglossal nucleus, 12N; trigeminal nucleus, 5N) as well as extensive portions of the IRN and parts of pons
Results, Brainstem activity maps of orofacial rhythms, paragraph 4; Figure 2gHow it will be judged
As fig2f-breathing-units-location, with licking-tuned units and the CCF regions XII, V and IRN, and the pons taken whole (P). Reproduced when the fraction exceeds that of all pons and medulla units in each of XII, V and IRN (Fisher exact, one-sided, p < 0.05) and licking units are present in the pons at more than 1% of its units.
- fig2h-swallow-units-location
Swallowing-related units clustered in the trigeminal nucleus (5N), lateral to the nucleus of the solitary tract (NTS) and above the nucleus ambiguus (10N), with further clusters in the inferior colliculus and the nucleus of the lateral lemniscus.
Units with swallowing-related activity were clustered in the trigeminal nucleus (5N), lateral to the nucleus of the solitary tract (NTS), and above the nucleus ambiguus (10N)
Results, Brainstem activity maps of orofacial rhythms, paragraph 4; Figure 2h; Extended Data Figure 6a-dHow it will be judged
As fig2f-breathing-units-location, with swallow-modulated units. Regions are CCF V, IC and NLL; "lateral to NTS" is units within (a) 300 um or (b) 500 um lateral of NTS at its AP extent; "above 10N" is units within the same distances dorsal of AMB. Reproduced when the fraction exceeds that of all units of the same sessions in all five regions under both distance readings; partially reproduced in three or four under both.
- ed6e-swallow-latencies
Swallow-related activity in 5N and IRN leads swallowing, while activity in the inferior colliculus and nucleus of the lateral lemniscus follows it.
Activities in 5N and IRN lead swallowing while activities in the IC and NLL follow swallowing.
Extended Data Figure 6e; Results, Brainstem activity maps of orofacial rhythms, paragraph 4How it will be judged
Swallow PSTHs as in the code (1 ms bins, 50 ms boxcar, -1 to 1 s). Per swallow-modulated unit, the time of the PSTH peak within -0.15 to 0.15 s (the paper gives no window or latency measure; a standard choice, named here). Reproduced when the median peak time is negative for units in V and IRN and positive for units in IC and NLL, and the two groups differ by a two-sided Mann-Whitney test at p < 0.05 (a standard choice, named here).
- ed7-glm-separate-clusters
A GLM on the tracked features found separate clusters of units predicted by licking or by breathing movements.
The GLM found separate clusters of units predicted by either licking or breathing movement (Extended Data Figure 7c-d)
Results, Brainstem activity maps of orofacial rhythms, paragraph 5; Extended Data Figure 7c-dHow it will be judged
The GLM of the Methods (17 ms bins, jaw and tongue x, y, z and airflow, z-scored, tongue set to 0 when its likelihood is below 0.05 in both views, every 5th trial held out, lags of +-5 bins), readings O8. The paper shows maps of the weights (Extended Data Fig 7c, airflow weights around and below 10N; 7d, jaw and tongue weights in the premotor regions) and gives no measure. Among units with validation VAF > 0.05, each unit is assigned to airflow or to jaw and tongue by its largest absolute weight (threshold and assignment are standard choices, named here). Reproduced when airflow units lie ventral of jaw and tongue units (CCF DV, one-sided Mann-Whitney test, p < 0.05) and the fraction of airflow units near the preBotzinger complex exceeds that of jaw and tongue units (one-sided Fisher exact test, p < 0.05), under every reading of O3 and O8; both tests are standard choices, named here.
- ed7g-cae-better-than-features
The autoencoder embedding predicted a larger fraction of activity variance than the tracked features for most units across the brainstem (Wilcoxon signed-rank test, P < 0.001).
The CAE predicted a larger fraction of activity variance than the tracked features for most units across the brainstem (Extended Data Figure 7g; CAE vs. features, P<0.001, Wilcoxon signed rank test)
Results, Brainstem activity maps of orofacial rhythms, paragraph 6; Extended Data Figure 7gHow it will be judged
CAE as in reference 40 and the Methods (two residual blocks of three convolutions, linear readouts of 128 and 32, frames downsampled to 120 x 112, one network per mouse trained on 80% of trials for 20,000 iterations with L2 regularization), which needs the raw videos; GLM readings O8. Reproduced when, over all units, CAE VAF exceeds feature VAF in more than half of the units and the paired Wilcoxon signed-rank test gives P < 0.001.
- ed7h-prebotc-exception
Units in the preBotzinger complex were the exception, where the autoencoder gave worse predictions than the tracked features (Wilcoxon signed-rank test, P < 0.001).
the only exceptions were units in the preBötzinger complex where activity prediction using CAE led to worse predictions (CAE vs. features, P<0.001, Wilcoxon signed rank test
Results, Brainstem activity maps of orofacial rhythms, paragraph 6; Extended Data Figure 7g-hHow it will be judged
As ed7g-cae-better-than-features, for units near the preBotzinger complex (O3 (a) and (b)), and for breathing-tuned units there (Extended Data Fig 7g marks breathing units). Reproduced when feature VAF exceeds CAE VAF with P < 0.001 under every reading of O3 and O8.
- ed8bc-prebotc-inspiration-phase
Breathing-related units in the preBotzinger complex preferentially fired at the inspiration phase of breathing.
units in the preBötzinger complex preferentially fired at inspiration phase of breathing
Results, Brainstem activity maps of orofacial rhythms, paragraph 5; Extended Data Figure 8a-cHow it will be judged
Preferred phases of the fitted tuning curves (header), breathing units by O1 (b) as Extended Data Fig 8b states, located by O3 (a) and (b). Inspiration phase is the half cycle starting at inspiration onset (phase 0 to pi in the convention where inspiration onset is 0). Reproduced when more than half of those units prefer that half cycle and their phases are non-uniform by a Rayleigh test at p < 0.05 (a standard choice, named here), under both readings of O3.
- ed8e-licking-units-protrusion
Licking-related units were mostly active when the tongue was fully protruded.
Units with licking-related activity were mostly active when the tongue was in the fully protruded state
Results, Brainstem activity maps of orofacial rhythms, paragraph 5; Extended Data Figure 8d-fHow it will be judged
Licking units by O1 (b). For each session, the jaw phase at the frame of maximal tongue protrusion in each lick (the frame of largest 3D tongue distance from its onset position; the paper does not define full protrusion, a standard choice named here) gives the protrusion phase; a unit's preferred phase is compared with its session's circular median protrusion phase. Reproduced when more than half of licking units prefer a phase within pi/2 of it (a standard choice, named here).
- ed8i-ipsilateral-lick-selectivity
Licking-related neurons showed selectivity for ipsilateral licking, while individual units within each hemisphere had diverse direction preferences, including the center direction.
Overall, licking-related neurons showed selectivity for ipsilateral licking (Extended Data Figure 8i)
Results, Brainstem activity maps of orofacial rhythms, paragraph 7; Extended Data Figure 8g-iHow it will be judged
Lick angle by O9 (a) and (b); spikes from -50 to 100 ms around each lick onset (Extended Data Fig 8g); left and right licks are those with angle below and above the midline. Per licking unit (O1 (a)), left-lick rate minus right-lick rate. Reproduced when this difference is larger for left-hemisphere than right-hemisphere units (CCF medial- lateral position against the midline) by a one-sided Mann-Whitney test at p < 0.05 (a standard choice, named here) and each hemisphere's median favours the ipsilateral side, under both readings of O9.
- ed4c-licking-map-insensitive-to-sorting
The licking activity map was insensitive to spike-sorting quality; raising the amplitude threshold from 150 to 500 uV left 2223 of 19,000 units and the same map.
Licking activity map was insensitive to the quality of spike sorting. To obtain units with the highest amplitude, the waveform amplitude quality metric was increased from 150 μV to 500 μV. This filtered the 19,000 units with amplitude > 150 μV to 2223 unit with amplitude > 500 μV.
Extended Data Figure 4c; Results, Brainstem activity maps of orofacial rhythms, paragraph 2How it will be judged
Units passing the filter of the header with amplitude > 500 uV. Reproduced when the count is within tolerance and, over CCF regions with at least 20 such units, the mean licking MI per region in this subset correlates with that in all units (Spearman, p < 0.05, positive; a standard choice, named here). The paper's 19,000 against 20,300 units is not resolved here; both are reported.
- fig3a-premotor-regions
Premotor neurons of the masseter and genioglossus occupied largely distinct regions of the pons and medulla, overlapping in the dorsal IRN.
Premotor neurons for the masseter and genioglossus muscles occupied largely distinct regions of the pons and medulla, with some overlap in the dorsal IRN (Figure 3a)
Results, A licking premotor network with autonomous rhythm-generation properties, paragraph 3; Figure 3a; Extended Data Figure 9a-bHow it will be judged
Not testable on the dandiset. From the premotor neuron positions in CCF, the density contours of Extended Data Fig 9a (genioglossus 0.0063 and masseter 0.0019 neurons per mm3); reproduced when the overlap of the two contours is under half of the smaller and lies in IRN dorsal to its DV midpoint (both standard choices, named here).
- fig3b-licking-units-in-premotor-regions
Units inside the tongue-jaw premotor regions showed significantly higher licking modulation than position-shuffled populations (Wilcoxon rank-sum test, P < 0.001).
Units inside these regions showed significantly higher licking modulation compared to position-shuffled populations (P < 0.001 Wilcoxon rank sum test)
Results, A licking premotor network with autonomous rhythm-generation properties, paragraph 3; Figure 3bHow it will be judged
Premotor regions are the union of the two density contours of Extended Data Fig 9a built from reference 43's premotor neurons. Position shuffling is not described; readings (a) licking MI of units inside against the MIs of the same number of units drawn at random from all units (positions shuffled across units); (b) inside against outside. Reproduced when the inside MIs are higher with rank-sum P < 0.001 under both readings.
- fig3ce-descending-stimulation-rhythmic-licking
Photostimulation of ALM (n = 11 mice) or latSC (n = 3 mice) elicited sustained rhythmic licking at the naturalistic 7 Hz frequency.
Photostimulation of ALM or latSC (20 or 40 Hz pulse trains for 2 or 1.3 s) elicited sustained rhythmic licking at the naturalistic 7 Hz frequency (Figure 3c and e)
Results, A licking premotor network with autonomous rhythm-generation properties, paragraph 4; Figure 3c and 3eHow it will be judged
Photostimulation trials of the ALM mice (cued licking task) and latSC mice. Lick rate as tongue onsets per second from the first lick to the end of the light, mean per mouse then across mice (the paper does not say how the rate is taken; a standard choice, named here). Reproduced when the mean across mice is within tolerance for each group and licking lasts to the end of the light in most trials (a lick in the last 300 ms, a standard choice, named here).
- fig3dfh-projections-and-rhythm-site
ALM and latSC axons targeted the contralateral brainstem and engulfed the tongue-jaw premotor regions; constant photostimulation of Phox2b+ neurons in the posterior-dorsal IRN/PARN elicited rhythmic licking, while anterior sites locked the jaw.
In contrast, photostimulating a posterior-dorsal region of IRN and the adjacent region in PARN (anterior-lateral to 12N) elicited rhythmic jaw movement and tongue protrusions
Results, A licking premotor network with autonomous rhythm-generation properties, paragraphs 4 and 5; Figure 3d, 3f, 3g and 3hHow it will be judged
Anatomy and fibre-site data, with no stated deposit. For the fibre sites (46 sites, 27 mice, Figure 3h), reproduced when the sites classed rhythmic lie posterior and dorsal of those not rhythmic (median CCF AP and DV positions) and within or next to IRN or PARN.
- ed10-site-classes
Of the brainstem photostimulation sites, 18 evoked rhythmic jaw movement with a peak at 7 Hz and rhythmic tongue protrusions, 11 evoked no rhythmic jaw movement and 4 evoked only a transient tongue protrusion.
How it will be judged
Jaw position Fourier spectrum during the 2 s of light (Extended Data Fig 10b). Sites are classed rhythmic when the spectrum has a peak between 3 and 15 Hz (the jaw band of the Methods) higher than in the 2 s before the light; classing rules beyond that are not given. The three classes sum to 33 of the 46 sites of Figure 3h; the paper does not say how the other 13 were classed, or whether the 4 transient sites are among the 11. Reproduced when the counts are within tolerance and the peak of the mean spectrum of rhythmic sites is within tolerance.
- fig4b-artificial-licking-keeps-lick-timing
During ALM-induced licking the lick-breath relationship of voluntary licking was recapitulated; lick timing anticipated the next breath at a fixed latency across breathing frequencies.
Notably, the same lick-breath temporal relationship was recapitulated during artificially induced licking. Lick timing anticipated the next breath and proceeded inspiration at a fixed latency across different breathing frequencies (Figure 4b)
Results, Coupled brainstem oscillators autonomously coordinate licking and breathing, paragraph 2; Figure 4bHow it will be judged
Cued licking task of the ALM mice. The Fig 1g measure (header) on photostimulation trials (licks during the 2 s light) and on Go-cue trials separately, pooled per mouse (Figure 4b, "across all mice"); session pooling as needed for the minimum of 40 cycles per half. Reproduced when the photostimulation difference is positive on average across mice by a one-sample two-tailed t-test at p < 0.05 (the paper's test of Fig 1g; the threshold is a standard choice, named here).
- fig4c-artificial-licking-breath-timing
During ALM-induced licking, inspiration timing was still influenced by future lick timing.
Conversely, inspiration timing was influenced by future lick timing (Figure 4c)
Results, Coupled brainstem oscillators autonomously coordinate licking and breathing, paragraph 2; Figure 4cHow it will be judged
As fig4b-artificial-licking-keeps-lick-timing, with the Fig 1i measure of the header.
- ed11b-alm-breathing-rate
In a subset of mice, ALM photostimulation also increased breathing rate.
In a subset of mice, ALM photostimulation also induced an increase in breathing rate (Extended Data Figure 11b)
Results, Coupled brainstem oscillators autonomously coordinate licking and breathing, paragraph 2; Extended Data Figure 11bHow it will be judged
Breathing rate (inspiration onsets per second) in the 2 s of light against the 1 s before, per photostimulation trial. Reproduced when at least one and fewer than all 11 mice show an increase by a paired two-sided Wilcoxon signed-rank test over trials at p < 0.05 (a standard choice, named here).
- fig4e-licking-units-anticipate-breath
Preceding a slower breath, the activity of licking neurons was slowed in anticipation of the next inspiration.
Preceding a slower breath, the activity of licking neurons was slowed in anticipation of the next inspiration (Figure 4e)
Results, Coupled brainstem oscillators autonomously coordinate licking and breathing, paragraph 3; Figure 4eHow it will be judged
The unit peak difference of the header (code get_peak_diff_lick_unit), slow minus fast, per licking unit, units by O1 and recordings by O2. The legend says "Same as b", the test of Fig 1g; one-sample two-tailed t-test over units at p < 0.05 (threshold a standard choice, named here). Reproduced when the mean is positive and significant under all readings; blocked as methods_underspecified when the readings disagree.
- fig4f-breathing-units-anticipate-lick
Preceding a slower lick, the activity of breathing neurons was slowed in anticipation of the upcoming lick.
Preceding a slower lick, the activity of breathing neurons was slowed in anticipation of the upcoming lick (Figure 4f)
Results, Coupled brainstem oscillators autonomously coordinate licking and breathing, paragraph 3; Figure 4fHow it will be judged
As fig4e-licking-units-anticipate-breath, with breathing units and the code's get_peak_diff_breath_unit.
- fig4d-simultaneous-recordings
Licking and breathing oscillators were recorded simultaneously in 43 recordings from 21 mice, with 1 to 28 licking-related and 1 to 12 breathing-related units per recording.
43 recordings in 21 mice; 1-28 licking-related units and 1-12 breathing-related units per recording
Results, Coupled brainstem oscillators autonomously coordinate licking and breathing, paragraph 3; Figure 4dHow it will be judged
Recordings with two probes, one with licking units (O1) in the posterior-dorsal IRN/PARN and one with breathing units near the preBotzinger complex (O3). Reproduced when the counts are within tolerance and the unit ranges per recording are within 1 to 28 and 1 to 12 widened by 10%, under every reading of O1 and O3.
- fig4g-symmetric-lags
The cross-correlation was temporally symmetric; breathing units lagged licking units for 57 pairs, licking units lagged breathing units for 34 pairs, and 12 pairs peaked at zero.
breathing units lagged licking units for 57 pairs, while other licking units lagged breathing units for 34 pairs, (the remaining 12 pairs had cross-correlation peak at 0)
Results, Coupled brainstem oscillators autonomously coordinate licking and breathing, paragraph 4How it will be judged
Significant pairs and lags by O6. Reproduced when the three counts are within tolerance under every reading of O6; partially reproduced when, under every reading, both lag directions each hold at least a quarter of significant pairs.
- ed12c-correlation-without-licking
Significant cross-correlation between licking and breathing unit pairs was observed even when mice were not licking.
Notably, significant cross-correlation between the unit pairs was observed even when mice were not licking, i.e. when the licking oscillator was inactive (Extended Data Figure 12c)
Results, Coupled brainstem oscillators autonomously coordinate licking and breathing, paragraph 4; Extended Data Figure 12cHow it will be judged
As fig4g-licking-breathing-pairs-correlated, on the baseline of the cued licking task (trial start to Go cue, no tongue detected; Extended Data Fig 12c). Reproduced when more than half of the pairs are significant under every reading of O6.
- fig4h-fewer-correlations-outside-oscillator
Units outside the licking oscillator were less often cross-correlated with breathing units (54 of 110 pairs), a significantly lower proportion than licking-breathing oscillator pairs (Chi-square test, p < 0.0001).
units outside of the licking oscillator exhibited less cross correlation with the breathing units (Figure 4h; 54/110 pairs with significant cross correlation; a significantly lower proportion compared to licking-breathing oscillator unit pairs, p<0.0001, Chi-square test)
Results, Coupled brainstem oscillators autonomously coordinate licking and breathing, paragraph 4; Figure 4hHow it will be judged
Non-licking units are not defined; readings (a) units of the licking-oscillator probe not licking-tuned (O1); (b) units of that probe outside the posterior-dorsal IRN/PARN (O3). Correlograms by O6. Chi-square test on the 2 x 2 table of significant and not significant pairs, without continuity correction (i) and with it (ii). Reproduced when the proportion is lower with p < 0.0001 under (a) or (b), for every reading of O6 and of the continuity correction; the counts are reported against the expected values.
- fig5b-voluntary-lick-latency
Mice initiated licking 127.6 +- 29.6 ms after the Go cue (mean +- std across mice, n = 11).
Mice initiated licking 127.6 ± 29.6 ms after the Go cue onset (defined as the onset of tongue appearance in the video, mean ± std across mice, n=11 mice
Results, Descending control patterns oscillators at movement initiation, paragraph 2; Figure 5a-bHow it will be judged
Cued licking task, Go-cue trials, first tongue onset after the cue within the 0.5 s response window; mean per mouse, then mean across the 11 mice.
- fig5c-preemptive-breath-adjustment
Before voluntary licking, breathing phase deviated from random 115 +- 19 ms before lick onset and became entrained to a phase that put expiration at the first lick (two-sample Kuiper test with Bonferroni correction, p < 0.0001).
115 ± 19 ms before licking onset (mean ± s.e.m., bootstrap), breathing phase began to deviate significantly from random distribution and became entrained to a specific phase (Figure 5c; p<0.0001, two sample Kuiper test with Bonferroni correction)
Results, Descending control patterns oscillators at movement initiation, paragraph 2; Figure 5cHow it will be judged
Go-cue trials of the cued licking task; breathing phase (Hilbert, header) aligned to the first lick; test and latency by O7 (a) and (b). Reproduced when, under both readings of O7, the latency is within tolerance and the phase at lick onset lies in expiration (phase pi to 2 pi).
- fig5e-photostim-lick-latency-longer
ALM photostimulation evoked licking with a latency of 230.7 +- 31.5 ms, significantly longer than voluntary licking (paired t-test, two-tailed, p < 0.001).
Photostimulation evoked licking with a latency of 230.7 ± 31.5 ms (Figure 5e), which was significantly longer than voluntarily initiated licking (p<0.001, paired t-test, two-tailed)
Results, Descending control patterns oscillators at movement initiation, paragraph 3; Figure 5eHow it will be judged
First tongue onset after light onset on photostimulation trials, mean per mouse. The paper does not give the unit of the paired test; readings (a) the 11 mice, against the Go-cue latencies of fig5b-voluntary-lick-latency; (b) sessions, Go-cue and photostimulation means of the same session. Reproduced when the mean across mice is within tolerance and the paired two-tailed t-test gives p < 0.001 under one reading.
- fig5f-artificial-licking-late-breath-adjustment
Artificially induced licking did not recapitulate the pre-emptive breath adjustment; breathing phase changed only 47.6 +- 10 ms before lick onset, a significantly shorter latency than before voluntary licking (bootstrap, P < 0.0001).
Breathing phase was altered only shortly before licking onset (Figure 5f; 47.6 ± 10 ms), a significantly shorter latency than the pre-emptive adjustment before voluntary licking (Figure 5g; P<0.0001, bootstrap)
Results, Descending control patterns oscillators at movement initiation, paragraph 3; Figure 5f (the Results also cite a Figure 5g that the legend does not describe)How it will be judged
As fig5c-preemptive-breath-adjustment on photostimulation trials, aligned to the first tongue protrusion after light onset. The bootstrap comparison is not described; the reading used is 10,000 resamples of trials with replacement, drawn within each condition, p the fraction where the induced latency is at least the voluntary one (a standard choice, named here; P < 0.0001 needs at least 10,000 resamples). Reproduced when, under both readings of O7, the latency is within tolerance and the comparison gives P < 0.0001.
Pipeline
- Paper
- done (1 of 1)
- Claims
- done
- Claims check
- done (ready)
- Attempt
- todo
- Questions
- none
- Verified
- none