Claims 29
From the paper Contextual gating of whisker-evoked responses by frontal cortex supports flexible decision making · claims file · checked 2026-10-21: ready
- fig1c-lick-probabilities
Mice lick in the response window on 78 +/- 2% of Go-tone Whisker trials and withhold licking in the other trial types (Go-tone No-Whisker 10 +/- 1%, Nogo-tone Whisker 13 +/- 2%, Nogo-tone No-Whisker 1 +/- 0.3%, No-tone Whisker 15 +/- 2%), mean +/- SEM over 35 recording sessions in 14 mice.
Mice learned to lick correctly in the reward-predicting trials including the Go-tone and whisker deflection (probability of licking correctly in the reporting window, P(lick) = 78 ± 2%, mean ± standard error of the mean, SEM, n = 35 sessions across N = 14 mice during electrophysiological recordings described later) ... i) Go-tone without whisker deflection, P(lick) = 10 ± 1%; ii) Nogo-tone with whisker deflection, P(lick) = 13 ± 2%; iii) Nogo-tone without whisker deflection, P(lick) = 1 ± 0.3%; and iv) No-tone with whisker deflection, P(lick) = 15 ± 2%
Results, A short-term memory context-dependent whisker detection task, paragraph 1; Figure 1cHow it will be judged
As code, Ghaderi2025_Figure1_c_Performance.m. Per recording session and trial type, the fraction of completed (non early-lick) trials with a lick in the response window (the NWB lick flag, first lick within 1 s after the whisker time, Methods Behavioral data analysis), all trials (no quiet selection); mean and SEM over the 35 sessions. Reproduced when all five means are within tolerance; partially reproduced when the Go-tone Whisker mean exceeds every other trial type's by at least 50 points but some value is out of tolerance.
- fig1f-movement-latencies
In quiet Hit trials the whisker deflection is followed by whisker protraction (time to 20% of maximum 58 +/- 8 ms), jaw opening (115 +/- 17 ms) and tongue protrusion (231 +/- 29 ms), and tongue-spout contact at 330 +/- 20 ms.
In quiet Hit trials, following the whisker deflection in Go-tone trials, mice rapidly initiated a sequence of movements starting with whisker protraction after ~30 ms (time to 20% max = 58 ± 8 ms), followed by jaw opening (115 ± 17 ms) and then protrusion of the tongue (231 ± 29 ms) (Fig. 1f ) resulting in tongue-spout contact triggering water reward delivery with latency 330 ± 20 ms ( n = 35 sessions during electrophysiological recordings)
Results, A short-term memory context-dependent whisker detection task, paragraph 3; Figure 1fHow it will be judged
Movement traces as code, Ghaderi2025_Figure1_f_Movements_Hit.m (quiet, completed Hit trials; whisker speed, jaw and tongue displacement in 10 ms bins, each trial minus its mean over 0.8 to 1.0 s, averaged per session). The lick latency as code, Ghaderi2025_Sup_Figure1_b_SpoutReactionTime.m (first lick time minus 1 s, licks outside 1 to 2 s dropped, mean per session over completed Hit trials, all trials; mean and SEM over sessions). Open detail, none of the code computes "time to 20% max". Readings (a) per session, the first bin after 1.0 s where the session-mean trace reaches 20% of its maximum over 1.0 to 1.2 s, then mean and SEM over sessions; (b) the same on each trial, averaged per session. Reproduced when, under a reading, the three movement latencies and the lick latency are within tolerance; partially reproduced when the order whisker, jaw, tongue, spout contact holds but a value is out of tolerance. Since the reading is a measure detail, the outcome is the one both readings give; if they differ, blocked as methods_underspecified.
- quiet-trials-half
Selecting quiet trials (no whisker or jaw movement in the last 200 ms of the delay) rejects about half of the recorded trials.
This precludes the involvement of orofacial motor or sensory reafference signals from contributing to our decoding analyses, but rejects around half of the recorded trials
Results, Decoding context from neuronal activity during the delay period, paragraph 5How it will be judged
The quiet rule of the header (median absolute deviation, Methods; the code's mean absolute deviation as variant). The fraction of trials failing it, pooled over the 35 sessions and also as the median over sessions. "Recorded trials" leaves the denominator open; readings (a) completed trials, (b) all trials including early-lick aborted ones, each judged by the same rule. Reproduced when the pooled fraction is within tolerance under a reading; partially reproduced when it lies between 30% and 70%. The tolerance is 10% of the value.
- fig2-auditory-window
Inactivation during the Auditory epoch reduces licking on Go-tone Whisker trials most for A1 (39 +/- 4%), and also for wM2 (33 +/- 6%), wS2 (26 +/- 4%) and wS1 (19 +/- 6%), while ALM and fpS1 inactivation causes no significant decrease.
For Go-tone Whisker trials, the largest impact of inactivation during the Auditory epoch was found for A1 (39 ± 4% reduction in lick probability, p = 0.0004, n = 17 sessions across N = 9 mice), but significant reductions were also found for inactivation of wM2 (33 ± 6% reduction in lick probability, p = 0.0004, n = 14 sessions across N = 9 mice), wS2 (26 ± 4% reduction in lick probability, p = 0.0004, n = 18 sessions across N = 10 mice) and wS1 (19 ± 6% reduction in lick probability, p = 0.002, n = 16 sessions across N = 10 mice). Importantly, inactivation of either ALM or fpS1 during the Auditory window did not cause any significant decrease in licking during the reporting period.
Results, Spatiotemporally-specific optogenetic inactivation, paragraph 3; Figure 2b-dHow it will be judged
The optogenetic test of code, Ghaderi2025_Figure2_d_Plick_Sessions.m, used by all three window claims. Per session, P(lick) over completed Go-tone Whisker trials with no light and with light in the window. The reduction is the mean over sessions of no-light minus light P(lick), in percentage points. Test (Methods and code), on paired sessions, draw 9 sessions with replacement (9 is the minimum session count over areas and windows), take the difference of means, p = (number of draws on the wrong side of zero + 1) / iterations, two-tailed as 2 min(p, 1 - p); Benjamini-Hochberg FDR over the six areas within each window, significant at 0.05. Iterations 1,000 (Methods) with 10,000 (code) as variant. Open details. (1) The plus-or-minus spread is SEM over sessions (Results, "Data are presented as mean ± SEM") or SD (Fig. 2d legend); both are reported, the value is judged on the mean. (2) The reduction may be relative (percent of the no-light P(lick)); readings (a) absolute points, (b) relative percent. Reproduced when under a reading of (2) the four means are within tolerance, the four areas are significant and ALM and fpS1 are not; partially reproduced when the pattern of significance holds and a value is out of tolerance, or when one area's significance differs.
- fig2-delay-window
Inactivation during the Delay epoch reduces Go-tone Whisker licking strongly for wM2 (45 +/- 7%) and ALM (32 +/- 5%), more modestly for wS1 (16 +/- 3%), and not significantly for A1, wS2 or fpS1.
Inactivation during the Delay epoch of Go-tone Whisker trials caused a significant and large reduction in lick probability during the reporting window for wM2 (45 ± 7% reduction in lick probability, p = 0.0004, n = 14 sessions across N = 9 mice) and ALM (32 ± 5% reduction in lick probability, p = 0.0004, n = 9 sessions across N = 8 mice). Inactivation of wS1 also appeared to significantly, albeit more modestly, reduce licking (16 ± 3% reduction in lick probability, p = 0.0004, n = 16 sessions across N = 10 mice). On the other hand, inactivation of A1, wS2 or fpS1 during the Delay epoch did not significantly change lick probability in the reporting window.
Results, Spatiotemporally-specific optogenetic inactivation, paragraph 3; Figure 2b-dHow it will be judged
As fig2-auditory-window, for the Delay window. Reproduced when under a reading the three means are within tolerance, wM2, ALM and wS1 are significant and A1, wS2 and fpS1 are not; partially reproduced as in fig2-auditory-window.
- fig2-whisker-window
Inactivation during the Whisker epoch reduces Go-tone Whisker licking for every area except fpS1, most for ALM (52 +/- 4%) and wM2 (48 +/- 4%), then wS1 (38 +/- 3%), wS2 (35 +/- 4%) and A1 (13 +/- 3%).
Finally, all tested cortical regions, except fpS1, showed a significant reduction in licking probability when inactivation occurred during the Whisker stimulus epoch. The largest effects were found for inactivation of ALM (52 ± 4% reduction in lick probability, p = 0.0003, n = 9 sessions across N = 8 mice) and wM2 (48 ± 4% reduction in lick probability, p = 0.0003, n = 14 sessions across N = 9 mice), followed by wS1 (38 ± 3% reduction in lick probability, p = 0.0003, n = 20 sessions across N = 10 mice), wS2 (35 ± 4% reduction in lick probability, p = 0.0003, n = 18 sessions across N = 10 mice), and A1 (13 ± 3% reduction in lick probability, p = 0.01, n = 17 sessions across N = 9 mice).
Results, Spatiotemporally-specific optogenetic inactivation, paragraph 3; Figure 2b-dHow it will be judged
As fig2-auditory-window, for the Whisker window. Reproduced when under a reading the five means are within tolerance, those five areas are significant and fpS1 is not; partially reproduced as in fig2-auditory-window.
- fig2-wm2-vs-alm-auditory
Inactivation during the auditory window reduces licking significantly more for wM2 than for ALM (33 +/- 6% against 8 +/- 5%).
The effect of inactivation during the auditory window was significantly stronger for wM2 than ALM (reduction in lick probabilities for wM2 = 33 ± 6 % vs ALM = 8 ± 5 %; bootstrap comparison, FDR-corrected, p = 0.009).
Results, Spatiotemporally-specific optogenetic inactivation, paragraph 3How it will be judged
Reductions as fig2-auditory-window. The code does not hold this comparison and the paper does not describe it beyond "bootstrap comparison, FDR-corrected". Readings (a) an unpaired bootstrap of the difference of the two areas' per-session reductions (sessions resampled with replacement within each area, 9 per area as in the Methods' subsampling), two-tailed as in the code's p rule; (b) the same with all sessions of each area resampled. The family for FDR is not given; the uncorrected p is reported with each reading, and FDR over the 15 area pairs of the Auditory window as a variant. Reproduced when both means are within tolerance and both readings give p < 0.05 with wM2 larger; blocked as methods_underspecified if the readings disagree.
- fig2-other-trial-types
Inactivation has little impact on licking in the trial types other than Go-tone Whisker.
whereas inactivation had little impact upon other trial types
Results, Spatiotemporally-specific optogenetic inactivation, paragraph 2How it will be judged
The finding is shown in Supplementary Fig. 3, which was not read; its code (Ghaderi2025_Sup_Figure3_a_Heatmaps.m, Sup_Figure3_b_Brainmaps.m) draws per-area, per-window changes in P(lick) for the other four trial types and runs no test. Measure as fig2-auditory-window for trial types 2 to 5 (6 areas, 3 windows, 4 trial types, 72 changes). "Little impact" is not defined. Readings (a) the bootstrap test of fig2-auditory-window, FDR within each window and trial type, finds no significant change in more than 7 of the 72 (10%); (b) no change exceeds 13 points, the smallest reduction the paper calls significant on Go-tone Whisker trials (A1, Whisker window). Reproduced when both readings hold; not reproduced when neither does; blocked as methods_underspecified when they differ. Every change and its p are reported.
- units-recorded
Across 35 sessions in 14 mice, 10,294 well-isolated units were recorded in the five areas (A1 780, wS1 1938, wS2 2064, wM2 2749, ALM 2763).
Altogether, across 35 sessions in 14 mice, we recorded 10,294 well-isolated units located in: i) A1 - 780 units across 14 sessions in 6 mice; ii) wS1 - 1938 units across 24 sessions in 12 mice; iii) wS2 - 2064 units across 24 sessions in 11 mice; iv) wM2 - 2749 units across 18 sessions in 8 mice; and v) ALM - 2763 units across 16 sessions in 6 mice.
Results, Spatiotemporal dynamics of neocortical activity, paragraph 1How it will be judged
Units the dataset marks as included (Bombcell single somatic units, well isolated), assigned to areas as in the header (probe target area and CCF label in the area's acronyms). The session and mouse counts per area (A1 14 and 6, wS1 24 and 12, wS2 24 and 11, wM2 18 and 8, ALM 16 and 6) are reported against the paper. Reproduced when all six unit counts are within tolerance; a count that matches only without the CCF filter is reported as such.
- fig3a-activity-maps
In quiet Hit trials, probe-averaged activity changes are largest in A1 in the first 30 ms after the tone, in wM2 and ALM in the last 200 ms of the delay, in wS1 and wS2 in the first 30 ms after the whisker deflection, and in frontal areas 300 to 500 ms after it.
Averaged across the first 30 ms after the onset of the tone, the auditory cue evoked a fast excitation of neurons primarily in A1. Averaged across the last 200 ms of the delay period, elevated neuronal activity appeared most prominent in frontal areas wM2 and ALM. In the 30 ms following whisker deflection, sensory-evoked activity was primarily localized to wS1 and wS2. The period 300-500 ms after whisker stimulation was used to assess licking-related neuronal activity, which was most prominent in frontal regions.
Results, Spatiotemporal dynamics of neocortical activity, paragraph 2; Figure 3aHow it will be judged
As Methods (Receiver operating characteristic analysis and activity maps) and the Fig. 3a legend. Per probe assigned to one of the five areas, the trial- and cell-averaged change in firing rate of all cortical neurons on the probe over quiet completed Hit trials, in the Auditory (0 to 0.03 s) and Delay (0.8 to 1.0 s) windows relative to the pre-auditory baseline, and in the Whisker (1.0 to 1.03 s) and Lick (1.3 to 1.5 s) windows relative to the pre-whisker baseline; per area, the mean over its probes. The baseline lengths are not given. Pre-auditory, -1 to 0 s, the baseline the Methods use elsewhere. Pre-whisker, readings (a) 0.95 to 1.0 s (the 50 ms of Fig. 7a), (b) 0.8 to 1.0 s; the outcome is the one both give, methods_underspecified when they differ. Reproduced when A1 has the largest area mean in the Auditory window, wM2 and ALM the two largest in the Delay window, wS1 and wS2 the two largest in the Whisker window, and wM2 and ALM the two largest in the Lick window; partially reproduced when three of the four hold. The paper reports no test; the per-probe values are shown.
- fig3b-delay-and-response-psth
In grand-average PSTHs of correct quiet trials, firing in wM2 and ALM is elevated during the delay of Go-tone trials compared with Nogo-tone trials, and after the whisker stimulus activity rises in Go-tone Whisker Hit trials compared with all other trial types in wS1, wS2, wM2 and ALM.
there were two obvious differences comparing Go-tone and Nogo-tone trials: i) there was a prominent elevation in firing rate in wM2 and ALM during the delay period of Go-tone trials compared to Nogo-tone trials; and ii) after the whisker stimulus there was a clear increase of activity in correct Go-tone Whisker trials compared to all other trial types in wS1, wS2, wM2 and ALM
Results, Spatiotemporal dynamics of neocortical activity, paragraph 3; Figure 3bHow it will be judged
PSTHs as code, Ghaderi2025_Figure3_b_PSTH_brainareas.m and Neuronal_data_10ms.m, per neuron, the mean over quiet, correct trials of each type, baseline (-1 to 0 s) subtracted. The figure carries no test, so the test is fixed from the paper's Statistics section (paired comparisons by Wilcoxon signed-rank over neurons). (i) For wM2 and ALM, the per-neuron mean over 0.8 to 1.0 s, Go-tone (types 1 and 2 pooled, and type 1 alone) against Nogo-tone (types 3 and 4 pooled, and type 3 alone); (ii) for wS1, wS2, wM2 and ALM, the mean over the paper's Licking window, 1.3 to 1.5 s (Results, Fig. 3c; Methods, activity maps), Hit against each of the four other correct trial types. The paper gives no window for "after the whisker stimulus"; it ties the increase to licking, so its Licking window is used, with 1.0 to 1.5 s shown as a variant. Reproduced when every comparison of (i) and (ii) is higher in the Go-tone (or Hit) condition with p < 0.05 under both pairings of (i); partially reproduced when (i) holds and (ii) fails for one area. The quiet rule follows the code of each figure (whisker speed only in Neuronal_data_10ms.m) as a variant.
- fig3c-delay-modulation-wm2-alm
In the last 200 ms of the delay of quiet Hit trials, more wM2 than ALM neurons are up-modulated relative to baseline (16.0% against 14.2%, p = 0.0499) and fewer are down-modulated (14.3% against 17.0%, p = 0.0062), by chi-squared tests.
The fraction of up-modulated neurons was slightly but significantly higher in wM2 (16.0%) than in ALM (14.2%) ( p = 0.0499, chi-squared test), whereas the fraction of down-modulated neurons was significantly lower in wM2 (14.3 %) compared to ALM (17.0 %) ( p = 0.0062, chi-squared test).
Results, Spatiotemporal dynamics of neocortical activity, paragraph 4; Figure 3cHow it will be judged
As code, Modulation.m and Ghaderi2025_Figure3_c_roc_hit.m. Per neuron, over quiet completed Hit trials, the spike count in 0.8 to 1.0 s against the count in -0.2 to 0 s of the same trials (two samples, labels window and baseline), AUC by perfcurve, selectivity index 2(AUC - 0.5); p is the fraction of 200 label permutations with |SI| at least the observed; up-modulated when SI > 0 and p < 0.05, down-modulated when SI < 0 and p < 0.05. Fractions over all neurons of the area; 2 x 2 Pearson chi-squared without continuity correction (MATLAB crosstab) for up against not-up and down against not-down. Reproduced when the four fractions are within tolerance and both tests give p < 0.05 in the stated directions; partially reproduced when both directions hold but a test or a value fails (the up-modulation p is near 0.05, so the result of each permutation seed is reported).
- fig3d-wm2-infragranular-delay
During the late delay of Go-tone Whisker trials, wM2 infragranular neurons change their firing more than supragranular neurons (1.2 +/- 0.1 Hz against 0.03 +/- 0.10 Hz, p = 1.7 x 10^-7, two-sample t-test), while no layer difference is detected in A1, wS2 or ALM.
Statistical tests indicated that neuronal activity during the late delay period of Go-tone trials differed significantly across cortical layers in wM2, with infragranular neurons having larger changes in firing rates than supragranular neurons (1.2 ± 0.1 Hz vs 0.03 ± 0.10 Hz; two-sample t-test, p = 1.7 × 10⁻⁷), whereas no significant differences were detected in A1, wS2, or ALM.
Results, Spatiotemporal dynamics of neocortical activity, paragraph 5; Figure 3dHow it will be judged
As code, Neuronal_data_10ms.m, Ghaderi2025_Figure3_d_PSTH_depth.ipynb and compute_layer_stats_pg.m. Per neuron, the mean 10 ms PSTH over quiet completed Hit trials (quiet by jaw and whisker, Methods; the code's whisker-speed-only rule here is a variant), minus its mean over -1 to 0 s, averaged over 0.8 to 1.0 s, in Hz. Layers from the CCF label, layer 1 and 2/3 supragranular, 4 granular, 5 and 6 infragranular. Unpaired equal-variance t-tests between the three layer pairs in each area. Reproduced when the two wM2 values are within tolerance, the wM2 infragranular against supragranular test gives p < 0.05, and no layer pair in A1, wS2 or ALM gives p < 0.05; partially reproduced when the wM2 result holds and one layer pair elsewhere is significant. The claim of no difference in A1, wS2 and ALM is judged by the same t-test at 0.05, uncorrected, as the code runs it.
- fig4-29-clusters
Clustering all 10,294 neurons on their average activity in the five correct trial types (PCA, spectral embedding, Gaussian mixture model) gives 29 clusters at the minimum of the Bayesian Information Criterion.
The BIC analysis indicated that 29 clusters provided the best trade-off between accuracy and parsimony.
Results, Spatiotemporal dynamics of neocortical activity, paragraph 6; Figure 4a; Methods, Clustering of neuronal responsesHow it will be judged
As Methods (Clustering of neuronal responses), with the code (Neuronal_data_100ms.m and Data_Clustering.ipynb) for what the Methods leave open. Per neuron, the mean over quiet correct trials of each of the five types in 10 ms bins over -1 to 2 s (300 bins, Methods), baseline (-1 to 0 s) subtracted, sessions with fewer than 3 trials of a type or 5 neurons skipped (code), the five PSTHs concatenated (1500 bins) and divided by their range; neurons mean-centred; 17 PCs (Methods; the code's shuffle rule, PCs whose variance exceeds the mean of 1000 row shuffles with Bonferroni p < 0.05/N, is shown); similarity exp(-d^2/2 sigma^2) on Euclidean distances with sigma tuned to a mean similarity of 0.5 (Methods); normalized Laplacian eigenvectors 2 to 17; diagonal-covariance GMM (1000 initializations, seed 0, code) for k in 10, 15, 20, 25 to 44, 45, 50, 55 (code); the k of lowest BIC. The code's version (100 ms bins, 150 per type, quiet by whisker speed only, exp(-d/0.07975)) is a variant and its k is reported. Reproduced when the BIC minimum is within tolerance (10% of 29, rounded); partially reproduced when it is between 20 and 38.
- fig4b-named-clusters
Some clusters represent sensory responses (cluster 1, fast auditory response, mostly A1 neurons; cluster 29, fast whisker response, mostly wS1 and wS2), clusters 11 and 25 fire during the delay only on Go-tone trials and consist mainly of wM2 and ALM neurons, and cluster 4 decreases its firing during the Go-tone delay and is spread over the five areas.
Cluster #1 had a large and fast response to the auditory cue and mostly consisted of neurons in A1, whereas Cluster #29 responded strongly and rapidly to whisker deflection and mostly consisted of neurons in wS1 and wS2. A few clusters showed strong context-dependent firing selectivity comparing Go-tone and Nogo-tone trials, most notably Clusters #11 and #25, which showed prominent delay period firing only in Go-tone trials and mainly consisted of neurons in wM2 and ALM ( Fig. 4b ). Interestingly, another cluster, Cluster #4, showed a prominent decrease in firing rate during the delay period selectively for Go-tone trials and neurons in this cluster were widely distributed across the five recorded cortical areas
Results, Spatiotemporal dynamics of neocortical activity, paragraph 7; Figure 4a-bHow it will be judged
Cluster membership from the paper's labels (Zenodo doi:10.5281/zenodo.17973874, processed_data/Data_Clustering.mat), linked to the dataset's units by the code's global unit index; the author adds the link claim. PSTHs as fig3b-delay-and-response-psth. Reproduced when (1) cluster 1 consists mostly of A1 neurons and has a mean response over 0 to 30 ms above its baseline; (2) cluster 29 consists mostly of wS1 and wS2 neurons and has a mean response over 1.0 to 1.03 s above baseline; (3) clusters 11 and 25 each consist mostly of wM2 and ALM neurons and have a higher mean over 0.8 to 1.0 s in Hit than in Nogo-tone Whisker CR trials; (4) cluster 4 has a lower mean over 0.8 to 1.0 s in Hit than in Nogo CR trials and at least 5% of its neurons in each area. The windows are the paper's Auditory, Whisker and Delay windows (Fig. 3c). The paper defines neither "mostly" nor "widely distributed". "Mostly" is open, with readings (a) more than half of the cluster's neurons, (b) the largest share of any area (for a pair, the two largest shares); the outcome is the one both give, and methods_underspecified when they differ. The 5% bound for "widely distributed" is the reader's and is a variant only; the fractions are reported. Partially reproduced when three of the four parts hold.
- fig5a-single-neuron-context
Comparing Hit with Nogo-tone Whisker correct rejection trials in the last 200 ms of the delay, wM2 and ALM have the highest proportions of context-modulated neurons (wM2, 20% increase firing by 8 Hz on average, 14% decrease by 3 Hz; ALM, 17% increase by 8 Hz, 16% decrease by 3 Hz), and wM2 has the largest proportion of Go-tone excited neurons.
Overall, we found the highest proportion of significantly modulated neurons in wM2 (20% of neurons increased firing on average by 8 Hz and 14% decreased firing by 3 Hz) and ALM (17% increased firing by 8 Hz and 16% decreased firing by 3 Hz). Thus, wM2 contained the largest proportion of Go-tone excited neurons during the delay period.
Results, Decoding context from neuronal activity during the delay period, paragraph 1; Figure 5aHow it will be judged
As code, Roc_hit_cr3.m and Ghaderi2025_Figure5_a_SingleNeuron.m. Per neuron, spike counts in 0.8 to 1.0 s on quiet completed Hit trials against Nogo-tone Whisker CR trials; AUC, SI and 200-permutation p as in fig3c-delay-modulation-wm2-alm; modulated when p < 0.05; the change in firing is the Hit minus CR mean rate (Hz). The two 3 Hz values are printed as whole numbers, so their tolerance is the rounding interval, 0.5 Hz, wider than 10% (0.3). Reproduced when the eight values are within tolerance, the wM2 up-modulated fraction is the largest of the five areas, and the total modulated fractions of wM2 and ALM each exceed those of A1, wS1 and wS2; partially reproduced when the wM2 up-modulated fraction is the largest and some value is out of tolerance.
- fig5b-context-decoding-delay
With 50 neurons per area, context (Hit against Nogo-tone Whisker CR) is decoded in the last 200 ms of the delay above baseline in every area, but far better in wM2 (64 +/- 1%) and ALM (61 +/- 1%) than in A1 (52 +/- 1%), wS1 (52 +/- 1%) and wS2 (52 +/- 0.4%).
Focusing on the last 200 ms of the delay period, during which orofacial movements are absent, the auditory context could be decoded from the firing of 50 neurons in each area with accuracy varying across cortical regions: i) A1, 52 ± 1%, p = 0.008, n = 8 sessions across N = 6 mice; ii) wS1, 52 ± 1%, p = 0.0003, n = 17 sessions across N = 9 mice; iii) wS2, 52 ± 0.4%, p = 0.0003, n = 19 sessions across N = 9 mice; iv) wM2, 64 ± 1%, p = 0.0002, n = 18 sessions across N = 8 mice; and v) ALM, 61 ± 1%, p = 0.0004, n = 16 sessions across N = 6 mice.
Results, Decoding context from neuronal activity during the delay period, paragraph 2; Figure 5bHow it will be judged
As Methods (Population decoding) and code, Decoding_NbNeurons5_50.m and Ghaderi2025_Figure5_b_Decoding.m. Area-sessions with at least 50 neurons; quiet completed Hit against Nogo-tone Whisker CR trials; per 10 ms bin a linear SVM on spike counts, majority class randomly undersampled, features z-scored, 5-fold stratified cross-validation, accuracy the mean over folds; 100 random subsets of 50 neurons (seed = repetition), averaged. Delay accuracy is the mean over bins in 0.8 to 1.0 s; baseline the mean over -1 to 0 s (Methods), with -0.8 to 0 s (code) as variant; two-sided Wilcoxon signed-rank over sessions. For the SVM, the Methods choose its parameters by grid search with nested cross-validation and give no grid; the code trains libsvm nu-SVC with default nu and no search. Since a grid would be the agent's choice and the parameter can change the outcome, both are run, (a) Methods, a linear C-SVM with C chosen by nested 5-fold search over 10^-3 to 10^3 in decades (a standard grid), (b) code; the outcome is the one both give, methods_underspecified when they differ. This applies to every claim that decodes as this one. Reproduced when the five accuracies are within tolerance, all five tests give p < 0.05, and wM2 and ALM each exceed A1, wS1 and wS2; partially reproduced when the ordering and the tests hold and a value is out of tolerance.
- fig5b-decoding-time-course
Context decoding is at chance in the baseline, highest in A1 right after the cue, and during the delay falls in A1, wS1 and wS2 but rises in wM2 and ALM; with fewer neurons accuracy drops but wM2 and ALM stay above A1, wS1 and wS2.
Decoding in the baseline period was at chance level (50%) for all cortical regions, as expected. Immediately, upon presentation of the auditory contextual cue, the decoding of Go vs Nogo trials was highest in A1, but context could also be decoded from all other recorded cortical regions. The ability to decode context during the delay period rapidly decreased in A1, wS1 and wS2, but increased for wM2 and ALM. ... Decreasing the number of neurons per cortical region through further subsampling reduced decoding accuracy but maintained the overall conclusion that wM2 and ALM more strongly encode context compared to A1, wS1 and wS2.
Results, Decoding context from neuronal activity during the delay period, paragraph 2; Figure 5b left and rightHow it will be judged
Decoding as fig5b-context-decoding-delay, for 5, 10, ..., 50 neurons. Reproduced when (1) the mean baseline accuracy (-1 to 0 s) of each area is within 50 +/- 5% (10% of the value; the paper gives no test of chance); (2) the mean accuracy over the early cue window is highest in A1; (3) for A1, wS1 and wS2 the mean over 0.8 to 1.0 s is lower than over the early cue window, and for wM2 and ALM it is higher; (4) at every neuron count run, the wM2 and ALM delay accuracies exceed those of A1, wS1 and wS2 (Fig. 5b legend, "independent of the number of considered neurons"). The early cue window is not given for decoding; readings (a) the paper's Auditory window, 0 to 0.03 s (Fig. 3c, Methods activity maps), (b) 0 to 0.2 s; the outcome is the one both give, methods_underspecified when they differ. The full time courses are shown. Partially reproduced when (1), (2) and (4) hold and (3) fails for one area.
- fig5c-cluster-removal
Removing cluster 11 reduces late-delay context decoding in wM2 (by 3.8 +/- 1.0%) and ALM (4.5 +/- 1.4%) with no significant effect in A1, wS1 and wS2; removing cluster 25 also reduces it in wM2 (1.6 +/- 0.6%) and ALM (2.0 +/- 0.6%); removing any other cluster does not significantly decrease decoding.
Removal of Cluster #11 reduced decoding accuracy by 3.8 ± 1.0% ( p = 0.008) for wM2 and by 4.5 ± 1.4% ( p = 0.008) for ALM, with no significant effect upon A1, wS1 and wS2. Removal of Cluster #25 also reduced significantly the decoding accuracy by 1.6 ± 0.6% ( p = 0.03) for wM2 and by 2.0 ± 0.6% ( p = 0.03) for ALM.
Results, Decoding context from neuronal activity during the delay period, paragraph 3; Figure 5cHow it will be judged
Cluster labels as fig4b-named-clusters. As code, Decoding_delay.m, Decoding_clusters_random_dropout.m and Ghaderi2025_Figure5_c_Dropcluster11.m. Decoding as fig5b-context-decoding-delay but with all neurons of each area-session (at least 5; 14 A1, 24 wS1, 19 wS2, 18 wM2, 16 ALM sessions), once with all and once without the cluster's neurons; the change is without minus with, averaged over 0.8 to 1.0 s; two-sided Wilcoxon signed-rank against zero over sessions, Benjamini-Hochberg FDR over the five areas for each cluster. Reproduced when the four decreases are within tolerance and significant, cluster 11 is not significant in A1, wS1 or wS2, and no cluster other than 11 and 25 gives a significant decrease in any area; partially reproduced when clusters 11 and 25 behave as stated and up to two other cluster-area pairs are significant.
- fig5d-null-space-decoding
Using only movement-null-space activity of all trials, context is decoded in the late delay much better in wM2 (74 +/- 2%) and ALM (71 +/- 3%) than in A1 (52 +/- 0.8%, not significant), wS1 (52 +/- 0.4%) and wS2 (53 +/- 0.5%).
During the late delay period, in the 200 ms before whisker stimulus and using only ‘null’ space neuronal activity, we were again able to decode the context with varying success across cortical regions: i) A1, 52 ± 0.8%, p = 0.09; ii) wS1, 52 ± 0.4%, p = 0.0003; iii) wS2, 53 ± 0.5%, p = 0.0004; iv) wM2, 74 ± 2%, p = 0.0002; and v) ALM, 71 ± 3%, p = 0.0004.
Results, Decoding context from neuronal activity during the delay period, paragraph 4; Figure 5dHow it will be judged
As Methods and code, Movement_subspace_Define.m, pg_decompose.m, Movement_subspace_Decoding_null.m and Ghaderi2025_Figure5_d_DecodingSubspaces.m. Area-sessions with more than 20 neurons; all completed trials (no quiet selection); spike counts in 10 ms bins z-scored per neuron over all bins and trials; a bin is movement when whisker speed exceeds the session's mean plus one SD over -1 to 0 s; orthogonal 10-dimensional potent and null subspaces from the movement and non-movement covariances (subspaceID orthogonal_subspaces, alpha 0, seed 101); decoding of Hit against Nogo-tone Whisker CR on the 10 null dimensions as fig5b-context-decoding-delay, without neuron subsampling; delay 0.8 to 1.0 s against baseline (-1 to 0 s Methods, -0.8 to 0 s code). Reproduced when the five accuracies are within tolerance, wS1, wS2, wM2 and ALM give p < 0.05 and A1 does not; partially reproduced when wM2 and ALM exceed 60% and every other area is below 56% but a value or a test differs.
- fig6a-pca-delay-states
In the first two PCs, A1, wS1 and wS2 activity returns near its start by the end of the delay, while in correct Go-tone trials wM2 and ALM settle within about 200 ms into a state far from the origin, reached with and without the whisker stimulus; Nogo correct rejections return to baseline and Miss trials fail to reach the Go-tone state. PC1 explains 11% (A1, wS1, wS2), 17% (wM2) and 14% (ALM) of variance, PC2 6% (A1, wS1, wS2) and 8% (wM2, ALM).
Neuronal activity in each cortical area exhibited a rapid response after tone onset in PC1 (explained variance: A1, 11%; wS1, 11%; wS2, 11%; wM2, 17%; and ALM 14%) and PC2 (explained variance: A1, 6%; wS1, 6%; wS2, 6%; wM2, 8%; and ALM 8%). Activity in A1, wS1 and wS2 followed an apparently cyclic trajectory, returning near to the start point by the end of the delay period. However, in correct Go-tone trials, neuronal activity in wM2 and ALM settled rapidly (within ~200 ms) into an apparent fixed point far from the origin ... Interestingly, in error trials where the mouse did not lick in response to the whisker stimulus preceded by a Go-tone (i.e., Miss trials), the neural activity in wM2 and ALM failed to reach the apparent attractor state of correct Go-tone trials.
Results, Context-dependent neuronal dynamics during the delay period, paragraph 1; Figure 6aHow it will be judged
As code, Ghaderi2025_Figure6_a_pca.m. Per area, a pseudo-population of all its neurons; per neuron, the mean over quiet completed trials of Hit, Nogo-tone Whisker CR, Go-tone No-Whisker CR and Miss in 50 ms bins over -1 to 2 s, baseline (-1 to 0 s) subtracted, the four conditions' first 60 bins concatenated, neurons with zero range dropped, each neuron divided by its range; PCA with time bins as observations; explained variance of PC1 and PC2. The legend's reading (PCs fitted on Hit and Nogo CR only, the other two projected) is a variant. The state at the end of the delay is the mean of PC1 and PC2 over 0.8 to 1.0 s; its distance from the baseline point (mean over -1 to 0 s) is compared across conditions. Reproduced when the ten values are within tolerance and, for wM2 and ALM, the end-of-delay distance of Hit and Go-tone No-Whisker CR each exceeds that of Nogo CR and Miss, while for A1, wS1 and wS2 the Hit distance at 1.0 s is below half its maximum over 0 to 1 s. Partially reproduced when the geometry holds and some explained variances are out of tolerance.
- fig6b-context-coding-direction
Projected on a context coding direction, late-delay activity in wM2 and ALM is high in Hit trials and significantly lower in Nogo-tone, Go-tone No-Whisker and Miss trials.
Activity in the context coding dimension was high in wM2 and ALM throughout the delay period in correct Go-tone Hit trials, but strongly reduced in the incorrect Go-tone Miss trials.
Results, Context-dependent neuronal dynamics during the delay period, paragraph 2; Figure 6bHow it will be judged
As Methods (Principal component analysis and coding directions) and code, Coding_direction_10ms.m and Ghaderi2025_Figure6_b_CD_context.m. Per area-session (at least 5 neurons), quiet completed trials in 10 ms bins; CD context is the mean over 0.8 to 1.0 s of the per-bin difference of Hit and Nogo CR mean spike-count vectors, from a random 20% of each class's trials (Methods; 30% in the code, a variant), unit length, Gram-Schmidt orthogonalized with the whisker and lick directions after it (the context direction first, so unchanged); the held-out trials are projected; per session, the mean projection over 0.8 to 1.0 s for Hit, Nogo CR, Go-tone No-Whisker CR and Miss; two-sided Wilcoxon signed-rank Hit against each, p times 3. Reproduced when in wM2 and ALM all three comparisons give corrected p < 0.05 with Hit higher; partially reproduced when two of three hold in each. The legend's p-values for the other areas are reported against the result, not judged.
- fig6c-temporal-correlation
In wM2 and ALM the Go-tone induces a long-lasting increase in correlated population activity during the delay in Hit and Go-tone No-Whisker trials, absent in Nogo-tone and Miss trials, while in A1, wS1 and wS2 the cue evokes only a brief increase.
Presentation of the auditory cue evoked a brief increase in correlated neuronal activity in A1, wS1 and wS2. However, in wM2 and ALM, the Go-tone induced a long-lasting increase in correlated neuronal activity during the delay period in both Hit and Go-tone No-Whisker Correct Rejection trials. In contrast, presentation of the Nogo-tone failed to induce network states with high temporal correlation. Importantly, the highly-correlated network state during the delay period was also absent in Miss trials
Results, Context-dependent neuronal dynamics during the delay period, paragraph 3; Figure 6cHow it will be judged
As Methods (Temporal correlation analysis) and code, Temporal_corrrelation_10ms.m and Ghaderi2025_Figure6_c_temporalCorrelation.m. Per area-session and quiet completed trial, the 300 x 300 Pearson correlation matrix of the population spike-count vectors of 10 ms bins over -1 to 2 s; averaged over trials of a type; minus the mean of the off-diagonal entries of the baseline block (-1 to 0 s); the mean over the delay block (0 to 1 s by 0 to 1 s, bins 100 to 200 as in the code) per session; Hit against Nogo CR, Go-tone No-Whisker CR and Miss by two-sided Wilcoxon signed-rank, p times 3. The code drops one A1 Hit session (the fourth) by hand; the result is given with and without it. The brief cue-evoked increase in A1, wS1 and wS2 is shown in the matrices and not judged, as the paper gives no measure of it. Reproduced when in wM2 and ALM all three comparisons give corrected p < 0.05 with Hit higher; partially reproduced when the Nogo and Miss comparisons hold in both areas. The legend's other p-values (wS1 Hit against Miss 0.006, wS2 Hit against Nogo 0.047) are reported, not judged.
- fig7b-early-whisker-response
Whisker-evoked activity in the Go-tone context is transiently reduced in wS1 and wS2 around 15 ms after the deflection and higher from about 30 ms on in most areas, compared with the Nogo-tone context.
we found significant changes in evoked activity, with an early transient reduction in the whisker response in wS1 and wS2 for Go-tone trials, followed by an overall increased activity in most cortical regions at later times
Results, Context-dependent processing of the whisker stimulus, paragraph 1; Figure 7bHow it will be judged
As code, psth_5ms.m and Ghaderi2025_Figure7_b_DeltaFR.m. Per neuron, the mean 5 ms PSTH over quiet completed Hit trials and over Nogo-tone Whisker CR trials, each minus its mean over 0.95 to 1.0 s, in Hz; the Hit minus CR difference per bin from 0.955 to 1.1 s; two-sided Wilcoxon signed-rank against zero over the area's neurons. FDR (Benjamini-Hochberg) is applied for each area as the legend says, with the code's FDR over all 5 areas x 30 bins as variant. Reproduced when wS1 and wS2 each have a significant negative bin between 1.010 and 1.020 s, and at least three of the five areas have significant positive bins from 1.030 s onward; partially reproduced when only one of wS1 and wS2 shows the early reduction.
- fig7c-no-whisker-cd-change
Activity projected on a whisker coding direction, defined from the first 30 ms of No-tone Whisker responses, shows no significant difference between Go-tone and Nogo-tone contexts.
However, no significant context differences were found in analyses of a whisker sensory response coding direction based on the 30 ms of activity evoked by whisker stimuli without a preceding tone cue and without subsequent licking
Results, Context-dependent processing of the whisker stimulus, paragraph 1; Figure 7cHow it will be judged
A claim of absence. As code, Coding_direction_5ms.m and Ghaderi2025_Figure7_c_StimCD.m. Per area-session (at least 5 neurons), 5 ms bins; CD whisker is the mean over 1.0 to 1.03 s minus the mean over -1 to 0 s of the No-tone Whisker CR trial-averaged population vector, from 30% of those trials, Gram-Schmidt orthogonalized to CD context and unit length; held-out quiet completed Hit and Nogo-tone Whisker CR trials projected; per session the mean projection in the bin at 1.03 s; the code's test, Kruskal-Wallis of Hit against CR over sessions followed by LSD comparisons, at 0.05. Reproduced when the test does not find a difference in any area; not reproduced when it does in some area. The difference and the paired Wilcoxon signed-rank p are given for each area either way.
- fig7d-context-decoding-after-whisker
Decoding context from baseline-subtracted activity in 5 ms bins after the whisker stimulus, significant decoding appears transiently in A1 and wS2 at 15 ms for one bin, and persistently in wM2 from 20 ms and in ALM from 25 ms after the stimulus.
we found that significant context-dependent sensory processing transiently appeared in A1 and wS2 at 15 ms after whisker stimulus for a single 5 ms-time bin; and appeared in a more lasting manner in wM2 at 20 ms after whisker stimulus (and all later time bins) and in ALM at 25 ms after whisker stimulus (and all later times)
Results, Context-dependent processing of the whisker stimulus, paragraph 2; Figure 7dHow it will be judged
As code, Decoding_afterwhisker_runX.m and Ghaderi2025_Figure7_d_ContextDecoding_WS.m. Per area-session (at least 5 neurons, all neurons), quiet completed Hit against Nogo-tone Whisker CR trials in 5 ms bins from 0.955 to 1.1 s; each trial's counts minus its count in one baseline bin, repeated for each of the 10 bins from 0.955 to 1.0 s and the accuracies averaged; SVM, balancing and folds as fig5b-context-decoding-delay; 5 runs (seeds 1 to 5) averaged. Per bin, accuracy against the mean over 0.955 to 1.0 s by two-sided Wilcoxon signed-rank over sessions, FDR over the area's bins (legend), with the code's FDR over all areas and bins as variant. Reproduced when the first bin of wM2 after which every bin up to 1.1 s is significant is at 1.015 to 1.025 s, the same for ALM is 5 to 10 ms later than wM2, and no such persistent run starts before 1.03 s in A1, wS1 or wS2; partially reproduced when wM2 precedes ALM but the onsets differ from the paper by more than one bin. The single-bin results in A1 and wS2 are reported.
- fig8a-wm2-lick-divergence
In the first 30 ms after the whisker stimulus, only wM2 activity diverges into separate states for lick and no-lick trials, mainly along the context coding direction, and at 30 ms distinguishes lick from no-lick trials both in correct Go-tone and error Nogo-tone conditions; ALM shows a weaker effect and A1, wS1 and wS2 none.
Only wM2 showed a clear divergence of trajectories into two distinct states corresponding to lick vs no-lick outcomes. This separation was primarily along the context coding direction. Single-session analyses confirmed that wM2 activity at 30 ms significantly distinguished lick from no-lick trials both in correct Go-tone and error Nogo-tone conditions ( Supplementary Fig. 17 ). ALM showed a weaker effect, and no significant differences were observed in A1, wS1, or wS2.
Results, Context-dependent processing of the whisker stimulus, paragraph 3; Figure 8aHow it will be judged
The test is in Supplementary Fig. 17, not read; its code (Ghaderi2025_Sup_Figure17_Context_CD_LickvsNolick.m with Coding_direction_5ms.m) fixes it. Per area-session, held-out quiet completed trials projected on CD context (5 ms bins, as fig6b-context-coding-direction at 5 ms) and averaged per trial group, Go-tone Whisker lick and no lick, Nogo-tone Whisker lick and no lick, No-tone Whisker lick and no lick; the value in the bin at 1.03 s; Kruskal-Wallis over the six groups across sessions, then LSD comparisons. Reproduced when in wM2 the Kruskal-Wallis p < 0.05 and both Go-tone lick against no lick and Nogo-tone lick against no lick are significant with lick higher, while A1, wS1 and wS2 show no significant lick against no lick comparison; partially reproduced when wM2 holds for one of the two contexts. ALM's comparisons are reported.
- fig8b-lick-coding-direction
At 30 ms after the whisker stimulus, activity along a lick coding direction defined from spontaneous licking separates lick from no-lick trials in wM2 for Go-tone, Nogo-tone (false alarm) and No-tone trials, in ALM only for correct trials, and not in A1, wS1 or wS2.
Projecting neuronal activity in the lick coding direction revealed separation between lick and no-lick trials for both correct and false alarm cases in wM2 at 30 ms after delivery of the whisker stimulus ( Fig. 8b ). The lick coding direction did not separate between lick and no-lick trials in A1, wS1 or wS2 at this early time point. Neuronal activity in ALM was higher in the lick coding direction only for correct trials, but false alarm lick trials were not different from no-lick trials at 30 ms post whisker stimulus time.
Results, Context-dependent processing of the whisker stimulus, paragraph 4; Figure 8bHow it will be judged
As Methods and code, Coding_direction_SpontLick.m and Ghaderi2025_Figure8_b_SpontLick_CD_LickvsNolick.m. Spontaneous licks are licks outside 0 to 6 s of any trial start, in bouts (lick mask, from Lick_times.mat) longer than 0.5 s, whose mean jaw position over 0 to 0.1 s after the lick exceeds 3 times its mean over -1 to 0 s; sessions with fewer than 2 such licks are dropped. CD lick is the mean over spontaneous licks and neurons' 5 ms counts in 0.2 to 0.5 s minus those in -1 to -0.5 s, unit length. Quiet completed trials are projected and averaged per session in six groups (Go-tone, Nogo-tone and No-tone Whisker, each lick and no lick); the value in the bin at 1.03 s; Kruskal-Wallis over the six groups across sessions, then Fisher LSD comparisons (MATLAB multcompare on the Kruskal-Wallis ranks). The legend's three p-values per context are read as that context's lick group against the three no-lick groups, which the paper does not state. Reproduced when in wM2 the Kruskal-Wallis p < 0.05 and each context's lick group exceeds that context's no-lick group significantly; in ALM the Go-tone lick group exceeds no lick and the Nogo-tone lick group does not; and A1, wS1 and wS2 show no significant lick against no-lick comparison. Partially reproduced when wM2 holds for two of the three contexts.
- figs18-later-lick-coding
Later, at 0.2 to 0.4 s after the whisker stimulus, the lick coding direction separates lick from no-lick trials in ALM, wM2, wS2 and wS1, with the effect largest in ALM and ranked in that order.
at later times, neuronal activity in the lick coding direction showed significant differentiation between lick and no-lick trials in ALM, wM2, wS2 and wS1 ranked according to the effect size quantified at 0.2-0.4 s after the whisker stimulus with ALM being largest
Results, Context-dependent processing of the whisker stimulus, paragraph 4 (Supplementary Fig. 18)How it will be judged
Supplementary Fig. 18 was not read; its code (Ghaderi2025_Sup_Figure18_SpontLick_CD_LickvsNolick_200mswindow.m) takes the projection of fig8b-lick-coding-direction averaged over 1.2 to 1.4 s with the same Kruskal-Wallis and LSD test. The effect size is not defined in the paper or the code. Readings (a) the difference of the mean projection of all lick groups and all no-lick groups divided by their pooled SD over sessions (Cohen's d); (b) the Kruskal-Wallis chi-squared statistic. Reproduced when ALM, wM2, wS2 and wS1 each give Kruskal-Wallis p < 0.05 with at least one lick group above its no-lick group, and under both readings ALM has the largest effect; the full ranking is reported. If the readings disagree on which area is largest, blocked as methods_underspecified.
Pipeline
- Paper
- done (1 of 1)
- Claims
- done
- Claims check
- done (ready)
- Attempt
- todo
- Questions
- none
- Verified
- none