{"id":"e622720c-6e61-43ad-b3d2-379a5db13b53","arxiv_id":"2507.09360","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":8,"one_line_summary":"Coordinated shifts in the half-activation voltages of a potassium current and an h-current in a pyloric CPG model maintain triphasic phase relations across a 60-fold range of cycle periods.","lead":"This computational neuroscience paper shows that a three-neuron model of the crustacean pyloric rhythm can keep the relative timing of its three bursts nearly constant while the cycle period ranges from 0.5 to 30 seconds, by smoothly shifting two ion-current activation parameters. It suggests a concrete cellular bookkeeping rule for how neuromodulation could preserve patterned motor output across speeds.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The proposed phase-maintenance mechanism requires sub-microvolt parameter precision near the Cornerstone bifurcation; without evidence that neuromodulation can achieve this, the biological claim is unsubstantiated.","rationale":"The reader's weakest assumption identifies precisely the load-bearing concern: the near-bifurcation 'ghost' mechanism requires parameter values within 1e-7 to 1e-10 V of the critical Cornerstone values. This is not a mere numerical inconvenience; it determines whether the central biological claim holds. The model's internal mathematics are standard—inverse-square-root scaling near saddle-node and blue-sky bifurcations is well established—and the reported simulations are self-consistent, so the concern is not about internal consistency but about external biological feasibility. The paper's stated goal is to describe how neuromodulation could adjust neuronal properties to preserve phase; if neuromodulation cannot position V½ with the required precision, the mechanisms would not operate over the demonstrated period range. The missing code and referenced supplementary tables strengthen the need for a conditional verdict, but the precision issue is the deeper scientific concern because it targets the causal mechanism itself. A noise-sensitivity experiment, as proposed, would directly test whether the phase-maintenance property survives realistic parameter scatter. Since the paper already identifies this as a modeling proof-of-principle and does not overclaim experimental confirmation, the appropriate verdict remains CONDITIONAL: the computational demonstration is coherent, but the biological applicability is unverified. No change to the reader's verdict is needed.","tokens_in":29436,"tokens_out":5399,"duration_ms":63408,"concrete_test":"Re-simulate the 20 parameter sets of Figure 8 after adding independent zero-mean Gaussian noise with σ = 1e-3 V (a typical neuromodulatory V½ shift) to every θK2 and θh, and measure each neuron's duty cycle and the LP/PY phase relative to AB. If phase deviations exceed roughly 10% or the triphasic order fails for periods above ~2 s, the proposed mechanism is not biologically plausible. Alternatively, compute the largest σ that preserves phase within 10% at 30 s; if that σ is below 1e-5 V, the required precision is unsupported.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's central claim is that coordinated variation of θK2 and θh can maintain triphasic phase across 0.5–30 s periods via inverse-square-root 'ghost' dynamics near the Cornerstone bifurcation. This requires the operating parameters to lie extremely close to critical values at long periods. For example, the Fig. 6 fit for PY in Ensemble 3 gives θh* = -0.041358046586 V, and the 30 s pattern in Fig. 8 uses θh = 0.04135804605 V, a difference of ~5e-10 V; the 20 s sets differ by ~1e-7 to 1e-10 V. Typical neuromodulatory shifts of voltage-of-half-activation are millivolt-scale (1e-3 V), so the mechanism's long-period regime demands precision 4–7 orders of magnitude finer than current experimental evidence supports. The paper provides no sensitivity or noise analysis, and the text reports failed searches only as 'data not shown' (Section 3.5). Since the abstract explicitly claims neuromodulation 'could adjust' neuronal properties to preserve phase, this precision requirement is load-bearing: if biological parameter control is coarse, the proposed mechanisms cannot operate as described at biologically relevant periods, even though the simulations are internally consistent. The sign inconsistency for θh across figure captions (positive in Figs. 2 and 8, negative in Figs. 5 and 6) is secondary but hampers verification.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper presents a three-neuron conductance-based model of a pyloric CPG motif, with an endogenously bursting driver (AB) and two follower neurons (LP, PY) whose isolated dynamics are bursting, silent, or spiking. The authors use their previously described Cornerstone codimension-2 bifurcation, controlled by the half-activation voltages θK2 and θh, to propose three cellular mechanisms: Mechanism 1 controls burst duration and interburst interval in bursting neurons, Mechanism 2 controls the duration of a pulse-triggered burst in silent neurons, and Mechanism 3 controls the duration of a pause in spiking neurons. They construct four ensembles with different follower activity regimes, select twenty parameter sets per ensemble to produce triphasic patterns with approximately one-third duty cycle at periods from 0.5 to 30 s, and report near-constant duty cycles and follower phases. Section 3.5 shows that, for Ensembles 1, 3, and 4, varying only one follower parameter in addition to the two AB parameters suffices to maintain triphasic rhythm, while Ensemble 2 could not be tuned this way. The paper also fits the observed durations to inverse-square-root laws and analyzes steady-state IK2 and Ih currents as a function of period.","tokens_in":29773,"tokens_out":4118,"duration_ms":48129,"significance":"If valid, the paper offers a biophysically interpretable family of mechanisms by which follower neurons with different intrinsic activity regimes could participate in phase-maintaining CPG patterns, and it makes falsifiable predictions: burst duration grows by adding stereotyped spikes, latency to firing in spiking neurons is controlled by Ih half-activation, and the relevant durations follow inverse-square-root scaling near bifurcation. Strengths include the explicit model equations and parameter tables, deterministic numerical integration with stated tolerances, quantified phase and duty-cycle ranges for all ensembles, and a clearly articulated relationship to the authors' earlier Cornerstone bifurcation framework. The main weakness is that the biological relevance of the mechanism is asserted without evidence that neurons can control the half-activation voltages with the precision used in the simulations; the paper also provides no sensitivity or noise analysis and reports no uncertainty for the curve fits.","major_comments":[{"comment":"The long-period regime of the proposed mechanisms requires parameter values extraordinarily close to the bifurcation: for example, the PY neuron in Ensemble 3 at 30 s uses θh = 0.04135804605 V while the fitted critical value is θh* = 0.041358046586 V (Fig. 6), a difference of about 5×10^-10 V, and other 20 s parameter sets differ by roughly 10^-7 to 10^-10 V. Since the abstract claims that neuromodulation 'could adjust' neuronal properties to preserve phase, this precision requirement is load-bearing. Typical neuromodulatory shifts of voltage-of-half-activation are millivolt-scale, four to seven orders of magnitude larger. No sensitivity analysis, noise analysis, or evidence of such fine parameter control is provided. I recommend adding an explicit robustness analysis: perturb each selected parameter set by realistic amounts (e.g., 0.1–1 mV and channel-noise-like fluctuations) and report the resulting phase variability, or clearly restrict the biological claim to the shorter-period range where precision requirements are less extreme.","section":"§3.3, §3.5, Figs. 2 and 8"},{"comment":"The near-constant duty cycle is partly constructed rather than emergent. The methods state that parameter sets were selected so that AB and each follower neuron would have approximately one-third duty cycle at each cycle period, and the twenty parameter sets were hand-selected to satisfy this criterion. Therefore the near-constant duty cycles in Figs. 3C, 4C, 5C, 6C, and 7D largely confirm the selection procedure rather than test the mechanism. The independent content is the maintenance of burst order and the near-constant phases near one-third and two-thirds; these are also targets of the selection, although the achievement of stable ordered triphasic patterns across all twenty periods is nontrivial. To strengthen the claim, the paper should demonstrate that the mechanism maintains phase for target duty cycles other than one-third, or show that small perturbations away from the selected parameter sets preserve the pattern; otherwise the phrase 'phase maintenance' should be qualified as 'phase maintenance under the specific construction used here.'","section":"§2.2 and §3.1–3.4"},{"comment":"The inverse-square-root law is a central quantitative claim, but the curve fits are reported without any measure of uncertainty or goodness of fit. The coefficients a, b, and θ* are given to many significant digits, yet no confidence intervals, residual plots, or χ² values are shown. Moreover, the fits are applied only to the twenty hand-selected successful parameter sets; the paper states in §3.5 that failed searches are 'data not shown,' and the four-parameter variation claim is also 'data not shown.' Because the selection of successful cases can bias the apparent scaling, the manuscript should provide the fitted exponent as a free parameter (rather than assuming the inverse-square-root form), together with residual diagnostics and at least a summary of the failed searches. Without this, the quantitative evidence for the inverse-square-root law is incomplete.","section":"§3.2–3.4, Figs. 5–7"},{"comment":"The sign convention for θh is inconsistent across the manuscript and must be unified. The model definition says θh is the negative value of the voltage of half-activation (V½h = -θh), and the Cornerstone value is stated as θh* = 0.041356548 V in §2.1. However, Figs. 5–7 use θh* = -0.041358046586 V and fix θh at -0.0415 V, while Figs. 2 and 8 list positive θh values such as 0.04135646145 V. Additionally, Fig. 6's caption says 'where θh is fixed at -0.0106999 V,' which appears to be a typo for θK2. These inconsistencies make reproduction difficult and obscure whether the long-period parameter sets are near the Cornerstone bifurcation in the same direction as claimed. The authors should adopt one convention, state it explicitly next to every table and figure, and correct the figure-caption typos.","section":"§2.1, Figs. 2, 5, 6, 7, 8"}],"minor_comments":[{"comment":"The abstract contains the typo 'We idescribe' and should read 'We describe.'","section":"Abstract"},{"comment":"The text refers to 'Fig. 4C' when describing driver and follower duty cycles for Ensemble 3; this should be Fig. 6C and Fig. 6D, respectively.","section":"§3.3"},{"comment":"The manuscript repeatedly cites 'supplementary Table 1' and 'supplementary Table 2' but these tables are not included in the submitted text. They should be provided, since the twenty parameter sets per ensemble are central to reproducibility.","section":"§3.5 and supplementary tables"},{"comment":"The table lists different reversal potentials for AB→LP in Ensemble 3 versus Ensemble 4, but the text does not explain why this difference is needed; a sentence motivating these choices would help.","section":"§2.1, Table 2"},{"comment":"The statement says the MATLAB code 'will be available' upon acceptance; for a computational paper, depositing the code in ModelDB before review would allow the referees to verify the simulations and curve fits.","section":"Data Availability"}],"recommendation":"major_revision","confidential_remarks":"The paper is internally consistent as a dynamical-systems simulation study, and the authors have been transparent about the model equations and parameter tables. The main risk is overclaiming biological relevance without a robustness analysis: the sub-microvolt precision required near the Cornerstone bifurcation is a correct concern that should be addressed either by adding noise/sensitivity tests or by carefully restricting the scope of the biological interpretation. I do not see grounds for rejection, because the framework is plausible and the reported patterns are concrete, but the load-bearing points above need to be fixed before the paper can be accepted."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nThe headline: this paper's real contribution is not the Cornerstone bifurcation or the three temporal-control mechanisms (those are from the group's 2014 paper) but the four-ensemble network demonstration and the observation that three parameters can suffice. The simulations are coherent, the equations and tables are given, and the reported near-constant phases across 0.5–30 s are internally consistent. I'd send this to a serious referee.\n\nWhat's good: the model is simple and reproducible in principle, the ensembles cover heterogeneous follower types (bursting, silent, spiking, mixed), and the three-parameter sufficiency result is a useful step toward thinking about how few things a neuromodulator must adjust. The comparison to experimental coregulation work (Khorkova & Golowasch; MacLean) is fair, and the discussion about network-imposed period bounds shows the authors know the limits of their mechanism.\n\nThe soft spots are real but addressable. First, the biological precision problem. To hit a 30 s period, the θh values in Fig. 8 differ from the critical corner by ~5e-10 V; even at 0.5 s the distance is ~1e-7 V. Neuromodulatory shifts in half-activation are millivolt-scale, so the ghost mechanism, as presented, requires precision four to seven orders of magnitude finer than anything known. The paper does not include a noise or sensitivity analysis, and the failed hand-searches are only 'data not shown.' This does not kill the theoretical mechanism, but it does mean the abstract's 'neuromodulation could adjust...' overstates what is demonstrated. A revision should add parameter precision analysis or soften the biological claim.\n\nSecond, the phase constancy is partly constructed: each neuron is explicitly tuned to one-third duty cycle at each period. That is acceptable for a sufficiency proof, but the paper should frame it as such and provide at least a local robustness measure, for instance how phase degrades as parameters are jittered. Third, no code or data are shipped, and the supplementary tables are missing from the preprint; the promise of ModelDB makes this fixable.\n\nMinor issues: the sign of θh is inconsistent across figures (positive in Figs 2 and 8, negative in Figs 5 and 6), and the curve fits lack confidence bounds.\n\nVerdict: worth a serious referee. The modeling is careful, the new network results are a genuine extension, and the precision concern can be addressed in revision. I would not cite it in its current form, but I would bring it to a reading group focused on CPG models or bifurcation-based mechanisms.","headline":"Solid model extension of the Cornerstone bifurcation, but the biological claim needs a serious noise analysis before it can be taken as a neuromodulatory mechanism.","tokens_in":30337,"tokens_out":3918,"would_cite":false,"duration_ms":43266,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["37N25","92C20","34C23"],"pacs":[],"model":"deepseek-v4-flash","headline":"By shifting the half-activation voltages of two ionic currents, a three-neuron pyloric CPG motif keeps a triphasic pattern across a 60-fold range of periods.","keywords":["phase maintenance","central pattern generator","pyloric network","Cornerstone bifurcation","blue-sky catastrophe","inverse-square-root law","neuromodulation","half-activation voltage"],"falsifier":"A dynamic-clamp or pharmacological experiment on an identified pyloric neuron that steps the half-activation voltage of $I_{K2}$ or $I_h$ toward its presumed critical value and measures burst duration, interburst interval, or post-inhibitory latency would settle it: the paper's claim predicts these quantities grow as $1/\\sqrt{|\\alpha-\\alpha^*|}$ without bound as $\\alpha$ approaches $\\alpha^*$, whereas a system with noisy or saturating channel regulation would show a floor or a finite maximum below that divergence.","tokens_in":29191,"feed_emoji":"🦞","tokens_out":14624,"duration_ms":147606,"temperature":0.7,"pith_summary":"Rhythmic behaviors such as digestion and locomotion must keep their internal timing while the overall speed changes, and this paper offers a biophysical explanation for how a small neural circuit can do this. Working with a three-neuron model of the crustacean pyloric pattern generator, it shows that shifting just two parameters — the half-activation voltages of a potassium current and a hyperpolarization-activated current — lengthens burst duration, interburst interval, or latency to fire in a controlled way. Because those lengths decide where each neuron's activity falls in the cycle, coordinating the two shifts keeps a triphasic pattern at near-constant phase while the period stretches from 0.5 to 30 seconds. The control is organized around the Cornerstone bifurcation, where the transitions between bursting, silence, and spiking meet; near that point, each controlled duration grows as the inverse square root of the parameter distance. If real neurons can regulate these voltages precisely enough, the same scheme would explain phase maintenance in networks whose follower neurons are naturally bursting, silent, or spiking.","feed_headline":"Two ion-channel knobs keep phase as cycle period grows 60-fold","feed_subtitle":"A three-neuron pyloric motif holds one-third duty cycles from 0.5 to 30 s by tuning two half-activation voltages","key_machinery":"The Cornerstone bifurcation is the organizing center: the codimension-2 point in the $(\\theta_{K2},\\theta_h)$ plane where the blue-sky catastrophe (the bursting-to-spiking transition) meets the saddle-node-on-invariant-circle bifurcation (the bursting-to-silence transition). Its ghost — the lingering slow passage near the bifurcation point — makes the interval controlled by each parameter scale as $1/\\sqrt{|\\alpha-\\alpha^*|}$, where $\\alpha^*$ is the critical parameter value. Because the two parameters act on complementary phases of the burst cycle, one on the active phase and one on the silent phase, moving them together along an arc toward the Cornerstone point lengthens the cycle period while preserving the ratio of burst duration to interburst interval.","core_discovery":"The paper's central claim is that the voltages of half-activation of two currents, $\\theta_{K2}$ for the non-inactivating potassium current and $\\theta_h$ for the hyperpolarization-activated current, form a two-parameter control space for the temporal structure of a pyloric CPG motif. Near the Cornerstone bifurcation — the codimension-2 point where a blue-sky catastrophe and a saddle-node on an invariant circle intersect — each endogenous activity regime provides one mechanism: in a bursting neuron, $\\theta_{K2}$ approaching its critical value lengthens the burst while $\\theta_h$ approaching its critical value lengthens the interburst interval; in a silent neuron, $\\theta_{K2}$ near the saddle-node for periodic orbits sets the duration of a single inhibition-triggered burst; and in a spiking neuron, $\\theta_h$ near the saddle-node for equilibria sets the duration of a pause after inhibition. In all three cases the controlled duration follows an inverse-square-root law, and the paper tunes these parameters along arcs toward the Cornerstone point to generate 20 cycle periods from 0.5 to 30 s while holding every neuron near one-third duty cycle. The result is shown in four ensembles — both followers bursting, both silent, both spiking, and one silent with one spiking — and in three of the four ensembles only three varying parameters (the driver's two plus one follower's one) are needed to recover the full range of triphasic patterns.","pith_inferences":["A biological reading the paper does not itself make: because dynamic-clamp experiments can shift $\\theta_h$ and $\\theta_{K2}$, the mechanism predicts that slowly ramping these parameters toward the Cornerstone values should produce diverging interburst intervals or latencies in real pyloric neurons, which would be a direct experimental test.","The extremely fine parameter tolerances (down to nanovolts) suggest that if this scheme operates in living neurons, it must be implemented by activity-dependent homeostatic regulation of channel kinetics rather than by fixed expression levels, something the model does not address.","The three-parameter success in Ensembles 1, 3, and 4 hints that larger circuits — or coupled CPGs such as the pyloric–gastric system — might maintain phase with a similarly low-dimensional control strategy, but the paper does not simulate such networks.","The inverse-square-root curve fits used parameter sets that were selected by hand to reach one-third duty cycles, so an open statistical question is how much parameter slack the pattern tolerates under natural variability; this determines whether the mechanism is plausible in real preparations."],"forward_implications":["If the Cornerstone mechanism is correct, phase maintenance requires no global coordinating signal: each neuron can independently set its own $\\theta_{K2}$ and $\\theta_h$ to fix its duty cycle at any period.","The inverse-square-root scaling is a fingerprint: recordings showing burst duration, interburst interval, or post-inhibitory latency growing as $1/\\sqrt{|\\alpha-\\alpha^*|}$ as a modulatory parameter is shifted would confirm the mechanism.","Across all ensembles the number of spikes per burst grows linearly with period at roughly 49 spikes per second, so longer bursts arise by adding stereotyped spikes, a signature that experimentalists can look for in phase-maintaining pyloric neurons.","The minimal-parameter result — three varying parameters in three of the four ensembles — implies the neuromodulatory load of phase maintenance can be small, needing only one follower parameter in addition to the pacemaker's two.","Because duty cycle is set by the ratio of the two parameter distances, the same mechanisms should support phase divisions other than one-third, generalizing the framework to rhythms with different phase structure."],"supporting_citations":[{"why":"Introduces the single-neuron model and the codimension-2 Cornerstone bifurcation that organizes the three activity regimes.","marker":"[48]"},{"why":"Establishes the blue-sky catastrophe transition from bursting to spiking and its inverse-square-root control of burst duration.","marker":"[49]"},{"why":"Provides the experimental pyloric phase-maintenance data showing burst duration and spikes per burst scale with cycle period.","marker":"[45]"},{"why":"Models the same tri-phasic pyloric motif and shows synaptic depression alone can support phase maintenance, serving as the comparison baseline.","marker":"[24]"},{"why":"Shows that shifting the half-activation voltage of slow conductances reproduces phase-maintaining delays in isolated pyloric neurons.","marker":"[27]"},{"why":"Demonstrates experimental coregulation of IA and Ih currents in rhythmically active neurons, motivating the two-current coregulation hypothesis.","marker":"[29]"},{"why":"Provides experimental evidence that neuromodulators coordinate ionic current expression to maintain functional motor patterns.","marker":"[59]"},{"why":"Documents ghost-bursting and inverse-square-root divergence for near-threshold burst behavior used to interpret the numerical fits.","marker":"[56]"}],"fun_headline_variants":["Two half-activation voltages lock phase over 60-fold periods","Pyloric CPG phase preserved by tuning two ion-channel voltages","Mechanisms from a Cornerstone bifurcation maintain triphasic timing","Inverse-square-root law links burst duration to channel voltage knobs","Three mechanisms keep CPG duty cycle constant across period ranges"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that biological neuromodulation can set and hold the half-activation voltages of the potassium and h-currents with the extreme precision used in the simulations, where some follower parameters sit less than a nanovolt from the Cornerstone values; if real channel kinetics are noisy or coarsely regulated, the inverse-square-root ghost that makes durations grow will not operate as described.","fun_headline_variants_meta":{"raw":{"variants":["Two half-activation voltages lock phase over 60-fold periods","Pyloric CPG phase preserved by tuning two ion-channel voltages","Mechanisms from a Cornerstone bifurcation maintain triphasic timing","Inverse-square-root law links burst duration to channel voltage knobs","Three mechanisms keep CPG duty cycle constant across period ranges"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00025,"raw_usage":{"total_tokens":1671,"prompt_tokens":1181,"completion_tokens":490,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":797,"completion_tokens_details":{"reasoning_tokens":415}},"tokens_in":797,"tokens_out":490,"duration_ms":6418,"temperature":1.0,"reasoning_tokens":415,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T17:58:21.142721+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A dynamic-clamp or pharmacological experiment on an identified pyloric neuron that steps the half-activation voltage of $I_{K2}$ or $I_h$ toward its presumed critical value and measures burst duration, interburst interval, or post-inhibitory latency would settle it: the paper's claim predicts these quantities grow as $1/\\sqrt{|\\alpha-\\alpha^*|}$ without bound as $\\alpha$ approaches $\\alpha^*$, whereas a system with noisy or saturating channel regulation would show a floor or a finite maximum below that divergence.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Introduces the single-neuron model and the codimension-2 Cornerstone bifurcation that organizes the three activity regimes."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Establishes the blue-sky catastrophe transition from bursting to spiking and its inverse-square-root control of burst duration."},{"cited_title":"Journal of Computational Neuroscience, 1997","cited_arxiv_id":null,"evidence_quote":"Provides the experimental pyloric phase-maintenance data showing burst duration and spikes per burst scale with cycle period."},{"cited_title":"Nadim, and A","cited_arxiv_id":null,"evidence_quote":"Models the same tri-phasic pyloric motif and shows synaptic depression alone can support phase maintenance, serving as the comparison baseline."},{"cited_title":"Journal of Neuroscience, 2009","cited_arxiv_id":null,"evidence_quote":"Shows that shifting the half-activation voltage of slow conductances reproduces phase-maintaining delays in isolated pyloric neurons."},{"cited_title":"Archila, and A.A","cited_arxiv_id":null,"evidence_quote":"Demonstrates experimental coregulation of IA and Ih currents in rhythmically active neurons, motivating the two-current coregulation hypothesis."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides experimental evidence that neuromodulators coordinate ionic current expression to maintain functional motor patterns."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Documents ghost-bursting and inverse-square-root divergence for near-threshold burst behavior used to interpret the numerical fits."}],"review_version":1}