Claims 29
From the paper Allen Institute - Locus coeruleus norepinephrine activity during dynamic foraging - Electrophysiology and fiber photometry · claims file · checked 2026-11-04: ready
- patchseq-sample
Whole-cell patch-clamp recordings were made from 96 LC-NE neurons in brain slices from 28 mice, retrogradely labelled from frontal cortex (27 cells, 8 mice), cerebellar cortex (15 cells, 7 mice) or spinal cord (54 cells, 13 mice).
we made whole-cell patch clamp recordings from LC-NE neurons in brain slices (96 cells, 28 mice). We retrogradely labelled neurons from frontal cortex (27 cells, 8 mice), cerebellar cortex (15 cells, 7 mice) or spinal cord (54 cells, 13 mice)
Results, LC-NE activity during action-outcome learning, paragraph 5How it will be judged
A cell is one recorded neuron (one ephys_roi_id, one NWB file, as the paper's code LCNE-patchseq-figures at 79fa20a reads them), a mouse is one donor, and the group is the injection target. The tolerance is zero because these are counts of the deposited cells themselves, and the paper's code reads exactly 96. Reproduced when every count matches; partially reproduced when the total is within 10% (87 to 106 cells) but a count differs. Recordings deposited twice are counted once (author, step 1.2). Cells that fail the Methods' inclusion rule (seal above 1 GOhm before break-in, initial access resistance below 20 MOhm and below 15% of input resistance; Methods, Patch-seq recordings) are reported where the files hold those values.
- patchseq-ne-identity
The RNA sequenced from each recorded cell's nucleus confirmed that every recorded cell was a noradrenergic neuron.
At the end of the recording, we extracted the nucleus for RNA sequencing 41 and confirmed the identity of each cell as a noradrenergic neuron.
Results, LC-NE activity during action-outcome learning, paragraph 5; Methods, cDNA and sequencingHow it will be judged
The paper does not say how identity was confirmed. Gene counts as CPM of exonic plus intronic reads (Methods, cDNA and sequencing). The markers the paper uses for noradrenergic identity elsewhere are Dbh, Th, Slc6a2 and Slc18a2 (Methods, Identification of noradrenergic cells; Retro-seq). Open, with these readings named before the analysis (a) Dbh detected (CPM above 0) in the cell; (b) all four markers detected; (c) the cell's transcriptomic type mapped to the LC-NE (Dbh) type, as the LCNE-patchseq-analysis diagram shows with MapMyCells. Reproduced when all cells with a transcriptome pass under a reading; partially reproduced when at least 90% pass. The fraction of the 96 cells with a transcriptome is reported.
- patchseq-ap-waveform-spinal-vs-cortex
The intracellular action potential waveforms of LC-NE neurons projecting to the spinal cord differ significantly from those of neurons projecting to isocortex (two-sided Welch's t-test, p < 0.01).
Neurons projecting to spinal cord had significantly different action potential waveforms than those projecting to isocortex or cerebellum.
Extended Data Fig. 11k, left; Results, LC-NE activity during action-outcome learning, paragraph 5How it will be judged
Method from the paper's code (LCNE-patchseq-figures at 79fa20a, code/spikes.py and code/generate_S14jk.py), as the legend names no waveform measure. For each cell, take the long-square rheobase sweeps (protocols X3LP_Rheo_DA_0 and X5LP_Rheo_DA_0) whose main pulse lasts at least 0.8 s, and choose the one of lowest absolute amplitude that has at least one spike during the pulse. Detect spike peaks as the voltage maximum between paired upward and downward crossings of -10 mV (after eFEL-style resampling at the sampling step), and average the raw voltage from -5 to 10 ms around every peak of that sweep. Min-max normalize each average over -2 to 4 ms, keep -3 to 6 ms, mean-centre across all cells, and take PC1 by SVD (its sign is arbitrary and does not affect the test). Compare PC1 between spinal cord and cortex cells with a two-sided Welch t-test at the cell level (scipy ttest_ind, equal_var=False). Reproduced when p < 0.01; partially reproduced when 0.01 <= p < 0.05. The group means, t, degrees of freedom and p are reported. The Methods and the code disagree on sweep inclusion. The Methods give a sweep inclusion rule (bridge balance below 20 MOhm and 15% of input resistance, bias current within 100 pA, RMS noise below 0.07 mV over 1.5 ms and 0.5 mV over 500 ms; Methods, Patch-seq recordings) that the code does not apply. Following protocol step 2, the main reading applies the Methods' sweep rule before choosing the rheobase sweep, where the files hold the values it needs; the variant is the code's, with no further sweep rule. Where the files lack those values the two coincide, and this is recorded. Results under both are shown.
- patchseq-ap-waveform-spinal-vs-cerebellum
The intracellular action potential waveforms of LC-NE neurons projecting to the spinal cord differ significantly from those of neurons projecting to cerebellum (two-sided Welch's t-test, p < 0.05).
Intracellular action potential waveforms (left; two-sided Welch’s \(t\) -tests, spinal cord vs. cortex, \(p < 0.01\) , spinal cord vs. cerebellum, \(p < 0.05\) )
Extended Data Fig. 11k, left; Results, LC-NE activity during action-outcome learning, paragraph 5How it will be judged
The same waveform PC1 as in patchseq-ap-waveform-spinal-vs-cortex, from the paper's code (LCNE-patchseq-figures at 79fa20a), with the same main reading (the Methods' sweep rule) and variant (the code's) of sweep inclusion. Compare PC1 between spinal cord and cerebellum cells with a two-sided Welch t-test at the cell level. Reproduced when p < 0.05; not reproduced otherwise. Group means, t, degrees of freedom and p are reported.
- patchseq-tau-cortex-vs-spinal
LC-NE neurons projecting to isocortex have significantly longer membrane time constants than neurons projecting to the spinal cord (two-sided Welch's t-test, p < 0.001).
Neurons projecting to isocortex had significantly longer membrane time constants than those projecting to cerebellum or spinal cord
Extended Data Fig. 11k, right; Results, LC-NE activity during action-outcome learning, paragraph 5How it will be judged
The Methods say the features were computed with the IPFX package (Methods, Patch-seq recordings). The paper's code (LCNE-patchseq-figures at 79fa20a) does not compute the time constant; it reads IPFX's tau from a frozen table, in ms, missing for 14 of 96 cells. Main reading, from the Methods and the dataset - compute tau per cell with IPFX's long-square subthreshold analysis at its default settings (exponential fits to the hyperpolarizing long-square responses, averaged over the qualifying sweeps), using the Methods' sweep inclusion rule where the files hold its values. Variant, from the code - the frozen values in code/data/LCNE_patchseq_S14_cell_table.csv at 79fa20a, keyed by ephys_roi_id (an outside input). Compare cortex and spinal cord cells with a two-sided Welch t-test at the cell level (code/generate_S14jk.py, scipy ttest_ind, equal_var=False). Reproduced when p < 0.001 and the cortex mean exceeds the spinal cord mean; partially reproduced when p < 0.05 with cortex longer. The number of cells with a time constant per group, the group means, t and p are reported for each reading.
- patchseq-tau-cortex-vs-cerebellum
LC-NE neurons projecting to isocortex have significantly longer membrane time constants than neurons projecting to cerebellum (two-sided Welch's t-test, p < 0.001).
membrane time constants (right; two-sided Welch’s \(t\) -tests, spinal cord vs. cortex, \(p < 0.001\) , cerebellum vs. cortex, \(p < 0.001\) ) showed differences by projection target
Extended Data Fig. 11k, right; Results, LC-NE activity during action-outcome learning, paragraph 5How it will be judged
The time constant as in patchseq-tau-cortex-vs-spinal (IPFX on the dataset, main; the code's frozen table, variant). Compare cortex and cerebellum cells with a two-sided Welch t-test at the cell level. Reproduced when p < 0.001 and the cortex mean exceeds the cerebellum mean; partially reproduced when p < 0.05 with cortex longer. Cells per group, means, t and p are reported for each reading.
- lc-definition
LC, defined in the CCF as the region at the 67th density percentile of pontine Dbh+ nuclei (34,998 cells, 8 mice), is a small dorsal region of 0.13 mm3 holding 1,507 +- 55 noradrenergic neurons per mouse, with 367 +- 95 more in sub-coeruleus.
We defined LC in the CCF as a small dorsal region (0.13 mm 3 in volume) containing approximately two-thirds of all pontine noradrenergic neurons (67th density percentile, 1,507 ± 55, mean ± s.d.).
Results, LC-NE projections are spatially organized, paragraph 1; Fig. 1a,bHow it will be judged
Method from Methods, Definition of LC-NE neurons - Dbh-Cre;H2B-GFP nuclei detected in SmartSPIM volumes, registered to CCFv3, restricted to caudal midbrain and dorsal pons, reflected into one hemisphere; local density by nearest neighbours; LC is the mesh at the 67th density percentile. Tolerances on counts are the paper's s.d.
- axon-density
LC-NE axons from one hemisphere reach a high density in all CNS regions except the striatum, at varying densities; projections to the telencephalon are mainly ipsilateral and those to other regions bilateral.
We found a high density of LC-NE axons in all CNS regions other than the striatum, with varying densities across regions (Fig. 1c–e and Extended Data Fig. 1 ). Projections to the telencephalon were predominantly ipsilateral
Results, LC-NE projections are spatially organized, paragraph 2; Fig. 1c-eHow it will be judged
From the unilateral dense labelling imaged by SPIM (Results, paragraph 2; Methods, Single-neuron reconstructions for the imaging). The paper gives no measure of density or a threshold, and does not say which set of regions the comparison spans. Open, with readings (a) the level-2 CCF regions, the level the paper uses for the reconstructions (Results, paragraph 3); (b) the coarse regions of Methods, Models of single-neuron axonal distributions, with striatum split out of cerebral nuclei. Reproduced when the striatum has the lowest axon density of the regions and the telencephalon's ipsilateral density exceeds its contralateral density.
- axon-length-and-branching
Reconstructed LC-NE neurons (132 cells, 4 mice) have long axons (34.20 +- 12.83 cm, longest 70.32 cm) that branch little (17.7 branches per cm), significantly less than 1,549 other CNS neurons (32.6 branches per cm; ANCOVA).
Many neurons had long axons (34.20 ± 12.83 cm, mean ± s.d.; Extended Data Fig. 2b ), with the largest neurons reaching 70.32 cm of total length ... LC-NE neurons branched significantly less (ANCOVA (analysis of covariance), P = 3.9 × 10 −160 )
Results, LC-NE projections are spatially organized, paragraph 3; Extended Data Fig. 2bHow it will be judged
Reconstructions in CCF from ExA-SPIM and HortaCloud (Methods, Single-neuron reconstructions); the comparison set is the MouseLight database. Reproduced when the values are within tolerance and the ANCOVA finds fewer branches per length for LC-NE neurons at p < 0.05.
- single-neuron-primary-targets
Individual LC-NE neurons tend to send their axons mainly to one region (cortex, thalamus and midbrain, cerebellum, brainstem or spinal cord); neurons projecting to the forebrain avoid the brainstem and spinal cord, and the reverse.
Consistent with largely non-overlapping pathways, neurons with forebrain projections tended to project to other regions in the forebrain but not the brainstem or spinal cord (Fig. 2c ).
Results, LC-NE projections are spatially organized, paragraph 3; Fig. 2b,cHow it will be judged
Fraction of each neuron's axon length in each coarse region and the rank correlation across neurons between region pairs (Fig. 2c legend; Methods, Models of single-neuron axonal distributions, for the regions). Reproduced when the rank correlations are positive among forebrain regions and negative between forebrain regions and medulla or spinal cord.
- markov-renewal-model
A model in which axons grow as a Markov random walk between brain regions with a length-dependent Weibull renewal process captures the anterior-posterior block structure of LC-NE projections, unlike random projections.
A simple model that assumes axons grow as a random walk with a renewal process that depends on the length of the axons captured fundamental structural features of LC-NE projections (Extended Data Fig. 2d ).
Results, LC-NE projections are spatially organized, paragraph 3; Extended Data Fig. 2d; Methods, Models of single-neuron axonal distributionsHow it will be judged
Model as in Methods, Models of single-neuron axonal distributions (10-um steps, transitions counted after 2 mm in the new region, minimum length 10 mm, Weibull renewal fit with incomplete spinal neurons censored). Reproduced when the model's region correlation matrix shows the anterior-posterior block structure of the data and the shuffled model does not.
- mapseq-consistent
MAPseq/BARseq barcoding finds LC-NE neurons projecting across the CNS, with soma and axon distributions consistent with the single-neuron reconstructions.
We found LC-NE neurons with projections across the CNS, with distributions of somata and axons consistent with our single-neuron morphologies (Extended Data Fig. 3 ).
Results, LC-NE projections are spatially organized, paragraph 5; Extended Data Fig. 3How it will be judged
Methods, BARseq and MAPseq and the following sections (1-mismatch barcode matching, 12 coarse regions, row-normalized fractions). The paper gives no measure of consistency. Reproduced when the share of neurons whose top region is forebrain against hindbrain follows the same dorsal-ventral soma order as in dorsal-ventral-projection-targets.
- dorsal-ventral-projection-targets
LC-NE neurons in dorsal LC have primary projections to the forebrain (olfactory bulb, isocortex, hippocampus) and neurons in ventral LC to the hindbrain (pons, medulla, spinal cord).
Neurons in more dorsal LC had primary projections to the forebrain (olfactory bulb, isocortex and hippocampus), whereas neurons in more ventral LC had primary projections to the hindbrain (pons, medulla and spinal cord).
Results, LC-NE projections are spatially organized, paragraph 6; Fig. 2d,eHow it will be judged
Soma CCF dorsal-ventral position of each reconstructed neuron against its primary target (region with the largest axon fraction). The paper reports no test. Reproduced when the median soma depth of forebrain-primary neurons is dorsal to that of hindbrain-primary neurons.
- retrograde-dorsal-cortex-ventral-spinal
Retrograde tracing confirms that noradrenergic neurons projecting to cortex lie preferentially in dorsal LC and those projecting to the spinal cord in ventral LC.
somata of noradrenergic neurons with projections to cortex were preferentially in dorsal LC and noradrenergic neurons with projections to the spinal cord were in ventral LC
Results, LC-NE projections are spatially organized, paragraph 6; Fig. 2f; Extended Data Fig. 5How it will be judged
AAVrg-DIO-FLPo in Dbh-Cre;Ai65 mice, labelled somata registered to CCF (Methods, Retrograde tracing). The paper reports no test here. Reproduced when the median dorsal-ventral CCF coordinate of cortex-labelled somata is dorsal to that of spinal cord-labelled somata.
- snrnaseq-single-cluster
LC-NE gene expression from snRNA-seq (4,743 Dbh+ neurons, 40 mice) is best described by a single cluster with smooth, graded variation (a continuum).
A single cluster best explained the data (Extended Data Fig. 6e ), although gene expression varied smoothly across the LC-NE neuron population.
Results, LC-NE neurons have graded gene expression in space, paragraph 1; Extended Data Fig. 6eHow it will be judged
Methods, Cluster quality analysis - silhouette scores for K = 1 to 15 with Gaussian mixture models in 10-dimensional PCA space (and MMIDAS), five seeds each. Reproduced when the mean silhouette score is highest at K = 1 under both clustering methods (Methods, Cluster quality analysis, "always highest with a single cluster").
- pseudospace-graded-genes
Many genes vary smoothly along the pseudospace axis of LC-NE gene expression, including axon guidance, adhesion, extracellular matrix, ion channel, receptor and transcription genes, some in opposing gradients (Epha6 against Epha7, Col18a1 against Col6a5).
In some cases, two related genes were expressed in opposing gradients ... Epha6 versus Epha7 and Col18a1 versus Col6a5
Results, LC-NE neurons have graded gene expression in space, paragraph 1; Fig. 3a,bHow it will be judged
Pseudospace as in Methods, Pseudospace analysis (cubic polynomial in the first 10 PCs of 1,500 HVGs after scVI, cells projected to the nearest of 10,000 curve points, scaled 0 to 1). Reproduced when Epha6 and Epha7, and Col18a1 and Col6a5, have rank correlations with pseudospace of opposite sign.
- merfish-pseudospace-dorsal-ventral
Pseudospace imputed onto 2,308 MERFISH LC-NE neurons registered to the CCF aligns with the dorsal-ventral axis of LC.
Gene expression pseudospace aligned with the dorsal–ventral axis of the LC (Fig. 3c and Extended Data Fig. 7 )
Results, LC-NE neurons have graded gene expression in space, paragraph 2; Fig. 3cHow it will be judged
Methods, Analysis of MERFISH data and Pseudospace score imputations (k = 50 nearest snRNA-seq neighbours on shared genes). Reproduced when imputed pseudospace correlates with the CCF dorsal-ventral coordinate at p < 0.05 and more strongly than with the other two axes.
- retroseq-projection-gene-space
Neurons projecting to isocortex, cerebellum or spinal cord occupy different parts of gene expression space; cortex-projecting neurons resemble dorsal LC-NE neurons, spinal cord-projecting neurons ventral ones, and cerebellum-projecting neurons lie between.
Cortex-projecting neurons shared gene expression with dorsal LC-NE neurons, spinal cord-projecting neurons with ventral neurons, and cerebellum-projecting neurons were in between.
Results, LC-NE neurons have graded gene expression in space, paragraph 3; Fig. 3dHow it will be judged
Projection targets imputed onto snRNA-seq cells by 10 nearest retro-seq neighbours with at least 8 agreeing (Methods, Projection site imputations for snRNA-seq data). Reproduced when pseudospace differs between imputed targets (Kruskal-Wallis p < 0.001) in the order cortex, cerebellum, spinal cord.
- projection-and-expression-axes
The primary spatial axes explaining axonal projections (retrograde tracing) and gene expression (MERFISH) both run dorsal-ventral, about 20 degrees apart, and differ significantly.
Primary axes that explain axonal projections (black, from retrograde tracing) and gene expression (grey, from MERFISH) were primarily dorsal–ventral but differed
Results, LC-NE neurons have graded gene expression in space, paragraph 4; Fig. 3eHow it will be judged
Methods, Global linear trend test for single features using LDA (retrograde labels), CCA (MERFISH genes), Bootstrap estimation of spatial vector confidence intervals and Comparison of two primary spatial axes. Reproduced when both axes are mainly dorsal-ventral, the angle is within tolerance and the bootstrap p < 0.05. Extended Data Fig. 15c prints p = 3.25 x 10^-7 for the same chi-squared; either p meets the criterion.
- behaviour-history-dependence
Mice tracked changing reward probabilities, using reward and choice history to guide choices; a logistic regression on outcome history fits choices with R2 = 0.60 and predicts 0.88 of them (259 sessions, 32 mice).
Mice tracked the changing reward probabilities, using reward and choice history to influence future choices (Fig. 4c ).
Fig. 4c; Results, LC-NE activity during action-outcome learning, paragraph 3How it will be judged
Choice model in Methods, Analyses and models of behaviour (10 trials of reward and choice history). Reproduced when both values are within tolerance and the reward history coefficients are positive and decay with lag.
- go-cue-response
Nearly all LC-NE neurons (240 of 241, 19 mice) increase their spike rate in the 300 ms after the go cue, and the size of this increase relative to baseline is anticorrelated with the correlation between go-cue and baseline rates.
Nearly all LC-NE neurons showed an increase in spike rate in a 300 ms window following the go cue (paired t -tests, 240 out of 241 cells, 19 mice; Fig. 4e–h ).
Results, LC-NE activity during action-outcome learning, paragraph 6; Fig. 4g,hHow it will be judged
Per cell, a paired t-test of the spike rate in the 300 ms after go cues against the 2 s before (Fig. 4g legend). Open, the paper does not give the threshold for counting a cell; readings (a) p < 0.05 with a higher rate after the cue; (b) a higher mean rate after the cue.
- ventral-respond-ignore
Spike rates in the 2 s before the go cue predict whether the mouse responds, higher before ignored trials (46.5% of neurons significant), strongest in ventral LC.
This effect was strongest in neurons in ventral LC (Fig. 5c ; P values from permutation test of linear regression between t -statistics and spatial location: R 2 = 0.152, P = 1.4 × 10 −3 ).
Results, LC-NE activity during action-outcome learning, paragraph 7; Fig. 5b-dHow it will be judged
Per neuron, linear regression of the baseline rate 2 s before each go cue on respond against ignore (Methods, Linear regression models of spike rates); spatial tests as in Methods, Analysing relationships with space (5,000 permutations; kNN with k = 20, 5 folds). Reproduced when the fraction is within tolerance and the linear permutation p < 0.05 with ventral neurons more positive for ignore.
- dorsal-switch-stay
LC-NE neurons with higher activity in the 500 ms after the go cue on switch than on stay choices lie in dorsal LC, and those preferring stay in ventral LC.
Neurons with higher activity during switch choices tended to be in dorsal LC, whereas those with higher activity during stay choices tended to be in ventral LC
Results, LC-NE activity during action-outcome learning, paragraph 8; Fig. 5f-hHow it will be judged
Per neuron, linear regression of the go-cue period rate (0.5 s) on switch against stay (Methods, Linear regression models of spike rates); spatial tests as in Methods, Analysing relationships with space. Reproduced when the linear permutation p < 0.05 with dorsal neurons more positive for switch.
- cortex-input-switch
LC-NE axonal fluorescence in prelimbic cortex and the spiking of LC-NE neurons identified as projecting to cortex are higher on switch than on stay choices, more so than the population of single neurons.
LC-NE axonal fluorescence changes in PL and single-neuron spiking projecting to frontal cortex were significantly higher during switch versus stay choices (Fig. 5j,k , Fisher’s exact test, P = 7.8 × 10 −5 ).
Results, LC-NE activity during action-outcome learning, paragraph 9; Fig. 5j,kHow it will be judged
Sign-based comparison of switch-versus-stay t-statistics (proportion positive) between groups, Fisher's exact test (Methods, Comparison between single-neuron activity and axonal GCaMP); axonal regression in 0.75 s after the go cue (Methods, Linear regression models of axonal GCaMP). Reproduced when p < 0.05 with a larger positive share for the cortex-projecting signals.
- outcome-responses-spatial
LC-NE neurons respond variously to reward or its absence (76.8% with significant regressions); reward-excited neurons lie more dorsal and anterior than no-reward-excited neurons, a nonuniform distribution in space.
Neurons with increased activity in response to reward tended to lie more dorsal and anterior in LC than neurons with increased activity in response to no reward
Results, Frontal cortex LC-NE activity correlates with RPE, paragraphs 1 and 2; Fig. 6aHow it will be judged
Outcome regressions as in Methods, Linear regression models of spike rates (reward and no-reward windows, the larger absolute t kept). The test behind P = 3.0 x 10^-4 is not named; the LDA trend test (Methods, Global linear trend test for single features using LDA) is the one the Methods give for categorical labels. Reproduced when the fraction is within tolerance and the spatial test gives p < 0.05 with reward-excited neurons dorsal.
- rpe-predicts-choice-change
The change in choice between consecutive trials correlates with model-estimated RPE (rank correlation 0.84).
Relationship between how choices changed on consecutive trials and RPE (rank correlation, ρ = 0.84, P = 0.0).
Fig. 6c; Results, Frontal cortex LC-NE activity correlates with RPE, paragraph 1How it will be judged
RPE from the hierarchical Q-learning model fit in Stan, change of choice likelihood as defined in Methods, Analyses and models of behaviour.
- outcome-coefficients-cluster
Regression coefficients of outcome-period spiking on reward and chosen value are nonuniformly distributed in angle, and neurons projecting to isocortex are biased towards RPE (excited by reward, inhibited by predicted reward).
Polar histogram of coefficients from regressions in e shows a nonuniform distribution of activity (Rayleigh test, z = 19.53, P = 2.3 × 10 −9 ).
Fig. 6e,f; Results, Frontal cortex LC-NE activity correlates with RPE, paragraph 2How it will be judged
Angle of the (R, Qc) coefficients per neuron (Methods, Linear regression models of spike rates); Mardia-Watson-Wheeler permutation test with 5,000 permutations (Methods, Circular distribution comparison). Reproduced when both p < 0.05.
- frontal-axons-rpe
LC-NE axonal activity in prelimbic cortex 1.5 s after outcome correlates with RPE, and the regressions of axonal activity and of cortex-projecting neurons are more positive than those of the population of single neurons.
Regression t -statistics (Fig. 6j ) and coefficients (Fig. 6k ) were more positive than the population of single neurons (Fisher’s exact test, P = 5.7 × 10 −3 ) and consistent with modulation by RPE.
Results, Frontal cortex LC-NE activity correlates with RPE, paragraph 3; Fig. 6i-lHow it will be judged
Axonal regressions in a 2 s window after outcome (Methods, Linear regression models of axonal GCaMP); comparisons as in Methods, Comparison between single-neuron activity and axonal GCaMP and Circular distribution comparison. Reproduced when both p < 0.05 with axonal coefficients towards positive reward and negative chosen value.
- extracellular-waveform-spatial
Extracellular spike waveforms of LC-NE neurons vary in space; dorsal neurons have more symmetric waveforms and longer peak-to-trough times than ventral neurons.
dorsal LC-NE neurons tended to have more symmetric spike waveforms and longer peak-to-trough times than ventral LC-NE neurons
Results, LC-NE activity during action-outcome learning, paragraph 4; Extended Data Fig. 11b-dHow it will be judged
Waveform features of optotagged units (Methods, Spike waveform extraction) and the spatial tests of Methods, Analysing relationships with space (kNN with k = 30 for waveforms). Reproduced when slope symmetry and PC1 both show a linear trend at p < 0.05 with dorsal neurons more symmetric.
Pipeline
- Paper
- done (1 of 1)
- Claims
- done
- Claims check
- done (ready)
- Attempt
- todo
- Questions
- none
- Verified
- none