A curvature-based criterion for harmonic circadian waveforms

Hiroshi Ito; Ryoya Fukasaku; Yusuke Yamada; Yutaro Kabata

REVIEW 2 major objections 4 minor 25 references

A curvature-based criterion for harmonic circadian waveforms

T0 review · 2 major / 4 minor · reviewed 2026-08-04 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read The paper defines a circadian waveform as harmonic when its phase-plane trajectory has no inflection point, and reports that cyanobacterial and mouse-SCN rhythms, most modeled clock variables, and every Goodwin-model limit cycle examined pa

desk verdict A genuinely new waveform criterion and a careful Goodwin-model analysis, but the empirical 'harmonic circadian waveforms' claim rests on a post-hoc smoothing choice that was not independently validated. read the letter →

arxiv 2608.02562 v1 pith:Q45GQEUT submitted 2026-08-03 q-bio.QM

Yusuke Yamada , Yutaro Kabata , Ryoya Fukasaku , Hiroshi Ito This is my paper

classification q-bio.QM

keywords circadianrhythmwaveforminflectionpointcurvatureharmonicoscillationGoodwinmodelphase-planetrajectoryGaussiansmoothing

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper aims to establish that circadian oscillations are geometrically smooth—harmonic in a phase-plane sense—rather than pulse-like. It defines a waveform as harmonic when the curve traced by (x, ẋ) never has an inflection point, and shows that real bioluminescence recordings from cyanobacteria and the mouse suprachiasmatic nucleus, the core variables of several circadian clock models, and the limit cycles of the minimal negative-feedback Goodwin model all satisfy this condition. The claim matters because waveform shape has been linked to clock function—photoperiod sensing, fat-cell differentiation, and the duration of rest—yet most analyses have focused on period and phase. The proposed criterion is model-free, requiring only a time series of one variable, so it offers a new way to classify biological oscillations generally.

What carries the argument

The central object is the curvature κ(t) = (ẋ ...x − ẍ²)/(ẋ² + ẍ²)^(3/2) along the phase-plane trajectory Γ(t) = (x(t), ẋ(t)). A waveform is classified harmonic when the curvature never changes sign, i.e. the trajectory never touches the condition ẋ d³x/dt³ = ẍ², which corresponds geometrically to an inflection point and locally to exponential-like behavior. The method detects inflection points as sign changes of κ, and uses Gaussian smoothing at a width chosen so that only aperiodic, noise-induced inflection pairs are annihilated while period-recurring ones survive.

What would settle it

Record a high-resolution circadian bioluminescence time series from cyanobacteria or the mouse SCN, apply curvature analysis either without smoothing or with a data-driven smoothing width, and check whether a sign change in κ recurs in every cycle; a single periodically recurring inflection point in the real data would falsify the central empirical claim. Likewise, solving the piecewise-linear limit-cycle equations to high precision and finding a λ for which the passage point violates inequality (16) would overturn the semi-analytic confirmation.

Watch

Extended reading notes

Core claim

Using the curvature of the trajectory in the plane of a variable and its time derivative, the authors define an oscillation as harmonic when no inflection point appears on the limit cycle. They report that bioluminescence rhythms of cyanobacteria and the mouse SCN, the core mRNA variables of a detailed mammalian circadian clock model, the mean KaiC phosphorylation rhythm of a stochastic cyanobacterial model, and the Goodwin negative-feedback model are all harmonic. For the Goodwin model, harmonicity holds throughout the two-parameter region that produces a limit cycle; a piecewise-linearized version confirms this semi-analytically, reducing the absence of inflection points to an explicit ine

Load-bearing premise

The empirical harmonicity claims rest on the assumption, stated in the Methods' noise-removal section, that Gaussian smoothing with width σ=0.1 removes exactly the aperiodic, noise-induced inflection points and preserves every real periodic one; the semi-analytic Goodwin result additionally relies, as noted in the piecewise-linear analysis section, on locating the limit cycle's crossing point numerically rather than proving that the inequality holds for all λ.

Editorial extensions

If this is right

Circadian waveforms can be characterized without fitting a dynamical model: one observed variable, its derivatives, and the sign of a single curvature expression suffice.
The minimal negative-feedback Goodwin model is harmonic for every parameter pair that yields a limit cycle, making harmonicity a generic property of that oscillator structure rather than a fine-tuned one.
Not every variable in a circadian model is harmonic—some species, such as one of the Cryptochrome mRNA forms, do exhibit inflection points—so harmonic class is a property of individual variables and may trace signal-processing stages.
In the piecewise-linear limit of the Goodwin model, the absence of inflection points is equivalent to an explicit inequality on the limit cycle's crossing point, giving a semi-analytic handle on waveform shape.
The same curvature criterion can be applied to any self-sustained biological rhythm, replacing a qualitative harmonic-versus-relaxation label with a computable, data-driven classification.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

An implicit consequence is that comparing harmonic classes across variables within a clock network could expose low-pass filtering steps: the paper notes a simple cascade attenuates high frequencies, so downstream species may show smoother waveforms than upstream ones.
The hourglass-shaped surface of inflection points in the Goodwin phase space suggests a geometric explanation for harmonicity; testing whether limit cycles in other feedback oscillators avoid analogous surfaces would connect waveform shape to oscillator topology.
Because the empirical result depends on a chosen Gaussian smoothing width, a data-driven procedure for selecting that width would extend the method to noisier recordings, such as human actigraphy or tissue-level bioluminescence.
Applied beyond chronobiology, the criterion offers a model-free way to separate spike-like (relaxation) oscillations from smooth ones in cardiac or neural data, where the same waveform question arises.

Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

Two things you should know. First, the paper defines a clean, model-free waveform class: an oscillation is 'harmonic' if its phase-plane trajectory has no inflection point, and this is genuinely new as a tool for classifying circadian waveforms. Second, the empirical headline — that real bioluminescence data from cyanobacteria and mouse SCN are harmonic — is not independently established, because the Gaussian smoothing width σ=0.1 was chosen after inspecting the data and was validated only on FitzHugh–Nagumo, where essential inflections are large. Weak periodic inflections in real data could be destroyed by that same smoothing.

Credit where due. The Goodwin-model part is the strongest. The numerical scan over n and λ (Fig. 4D) is careful, and the piecewise-linear sufficient condition (Eq. 16) is derived correctly. The phase-space picture of the inflection surface (Fig. 6) is a nice geometric addition. The paper is also honest: the Discussion concedes that smoothing strength affects inflection detection and that statistical methods are needed.

Soft spots, in order of severity. The smoothing issue is load-bearing. Manual digitization of published figures at 0.5 h gives only 48 points per cycle; a third derivative from that is extremely noisy, so the result is essentially determined by the smoothing pipeline. The validation on synthetic data with spike-like inflections does not transfer to mild non-harmonicity. This alone means the empirical conclusion should be read as provisional. Second, the 'semi-analytic' confirmation of the Goodwin model covers only the piecewise-linear z–ż plane for a few numerically obtained passage points; the claim of parameter-independent harmonicity is supported mainly by the numerical scan, not by proof. That is fine if stated as a numerical result, but the word 'throughout' overstates it. Both issues are fixable.

Who is this for? Chronobiologists who care about waveform, and any dynamical-systems person interested in a threshold-free curvature-based classification. It deserves a serious referee; the idea is simple, checkable, and likely to be cited if the empirical part is tightened. My recommendation: engage with it, but send it back for revision focused on a principled smoothing-robustness analysis and softer wording of the empirical claim.

Referee Report

2 major / 4 minor

Summary. This paper introduces a geometric criterion for classifying oscillatory waveforms: a periodic time series is called 'harmonic' when its trajectory in the (x, ẋ) plane has no inflection point (curvature sign change). The authors derive the curvature condition, illustrate it on Stuart–Landau (harmonic) and FitzHugh–Nagumo (non-harmonic), and apply the criterion to experimental bioluminescence data from cyanobacteria and the mouse SCN, to two published circadian models (Sasai stochastic KaiC, Kim–Forger), and to the Goodwin model. For the Goodwin model they report a numerical parameter scan (n–λ) with no inflections wherever a limit cycle exists, and they give a partial semi-analytic confirmation using a piecewise-linear n→∞ version, checking a sufficient condition for the passage point.

Significance. Should the result hold, the paper offers a useful model-free descriptor of waveform shape, with potential for classifying circadian oscillators and for probing links between network architecture and waveform. Strengths: the numerical scan in Fig. 4D is carefully executed; arbitrary-precision arithmetic is used where ordinary double precision fails; the sufficient condition (Eq. 16) is derived correctly; and code is made available on GitHub. The principal weakness is the load-bearing choice of Gaussian smoothing width for the experimental data: σ=0.1 is selected after inspection and validated only on FitzHugh–Nagumo-type large inflections, so the conclusion that bioluminescence rhythms are harmonic may be an artifact of the denoising pipeline. The semi-analytic Goodwin result is also narrower than the abstract suggests, applying to the z–ż plane for a piecewise-linear limit and relying on numerically located passage points.

major comments (2)

[Methods, 'Noise removal from experimental data'; Fig. 2 and Fig. S2] The experimental harmonicity claim in Fig. 2 is not robustly established because the Gaussian smoothing width σ=0.1 is selected after inspecting the data ('applying σ=0.1 eliminated the aperiodic inflection points'), and the validation on FitzHugh–Nagumo does not transfer to circadian waveforms: the essential inflections there are large and spike-like, whereas mildly non-harmonic circadian rhythms would have weak inflections that the same smoothing can annihilate. The records are short (~6 cycles), and the paper concedes that smoothing strength can affect detection. Please report a sensitivity analysis of the inflection-point outcome over σ for the real data and validate the protocol on synthetic circadian-like signals with weak periodic inflections; otherwise the empirical conclusion may be an artifact of the pipeline.
[Harmonicity of the piecewise-linearized Goodwin model; Methods, Eq. (16)–(20)] The semi-analytical confirmation is more limited than the abstract's phrase 'confirmed this numerical trend semi-analytically' implies. Condition (16) is a sufficient condition for the absence of inflections for all t, but the passage point of the limit cycle is obtained by numerically solving the simultaneous equations (17)–(18), and the check is performed only for several values of λ. In addition, the proof concerns the z–ż plane of the n→∞ piecewise-linear model, whereas the numerical claim in Fig. 4D concerns the x–ẋ plane of the finite-n Goodwin model. Thus the analytical result does not by itself establish harmonicity throughout the n–λ region. Please state these restrictions explicitly in the abstract and in the section's summary, and avoid the suggestion that the full numerical scan has been proved analytically.

minor comments (4)

[Methods, 'Numerical computation of the curvature'] The sentence 'the Python packagempmath' is missing a space: it should read 'the Python package mpmath'.
[Results, 'Classification of waveforms from circadian clock models'] The gene name 'Bmals' is non-standard; for the mouse, the usual name is 'Bmal1'. Please revise or clarify.
[Data availability] The experimental time series were 'collected manually from the published figures' without specifying which figures in the cited papers, and the extracted series are not deposited. Please include the digitized data in the GitHub repository so that Fig. 2 can be reproduced.
[Introduction] The term 'harmonic' is already used in Fourier analysis; consider a short remark distinguishing the new geometric sense from the spectral sense to avoid ambiguity.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the curvature criterion is applied to independent data and models, and the few parameter choices (e.g., smoothing strength) are openly acknowledged limitations rather than fitted inputs renamed as predictions.

full rationale

The paper's central move is to define harmonicity geometrically (no inflection point in the (x, x-dot) plane) and then apply that definition to experimental time series and model outputs. The Goodwin-model claim is a numerical scan over the n-lambda limit-cycle region: 'no inflection point was observed for any parameter pair that produced a limit cycle' (Results, 'Harmonicity of the Goodwin model'). No parameter is fitted to force harmonicity; the criterion is evaluated on trajectories obtained by direct integration. The piecewise-linear analysis is explicitly semi-analytic: the authors derive a sufficient condition (Eq. 16) and then check numerically that the limit-cycle passage point satisfies it, noting that condition (i) 'provides a sufficient condition' and 'imposes a stronger constraint than necessary.' This is a numerical verification of the model's own equations, not a circular reduction. For the experimental data, the paper applies Gaussian smoothing and transparently states the limitation: 'The choice of smoothing strength can affect the detection of inflection points. Here we adopted a smoothing strength sufficient to remove those inflection points judged to be non-essential.' The smoothing parameter is chosen by examining periodicity of inflection points and validated on independent stochastic FitzHugh-Nagumo and Stuart-Landau models; it is not a fitted parameter silently renamed as a prediction. Self-citations (e.g., Kaji et al. 2025) appear only as background motivation and are not load-bearing for the geometric or Goodwin results. The remaining concerns about smoothing strength are robustness/correctness issues, not circularity.

Assumptions & free parameters 2 free parameters · 4 assumptions · 0 invented entities

The central claims rest on the Goodwin model as a circadian proxy, the piecewise-linear simplification for analytic tractability, and the Gaussian-smoothing preprocessing that determines the experimental verdict. These are the main unproven premises; beyond them, no fitted constants are used to force harmonicity.

free parameters (2)

Gaussian smoothing width σ (experimental data) = 0.1
Hand-chosen so that aperiodic inflection points disappear; directly determines whether data are classified as harmonic.
Noise intensities D in synthetic validation = 0.0016 (FN), 0.09 (SL)
Used to calibrate σ; not fitted to circadian data, but the choice affects the validation of smoothing.

assumptions (4)

domain assumption The Goodwin model (Eq. 5) is a minimal representation of the core circadian oscillator
Bridges model harmonicity to biological circadian clocks (Results: Harmonicity of the Goodwin model).
ad hoc to paper Piecewise-linear limit n→∞ preserves the inflection-point property of the finite-n Goodwin model
Semi-analytic proof is for this simplified model; no analytic link to finite n is established.
ad hoc to paper Gaussian smoothing with σ=0.1 removes only non-essential inflection points from real data
Calibrated on synthetic FitzHugh–Nagumo data; applied post hoc to cyanobacterial and SCN data (Methods).
domain assumption Limit cycle exists for the parameter region in Eq. (7)
Invokes Griffith (1968) linear stability analysis; parameter scan is restricted to oscillatory pairs.

how reviews work

0 comments

Cite this review

Pith. "Pith review of A curvature-based criterion for harmonic circadian waveforms." pith.science (2026). https://pith.science/paper/Q45GQEUT

@misc{pith2026260802562,
  author       = {Pith},
  title        = {Pith review of: A curvature-based criterion for harmonic circadian waveforms},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/Q45GQEUT}},
  note         = {Machine review of arXiv:2608.02562}
}

read the original abstract

Experimental and theoretical studies of circadian rhythms have focused largely on the period, on mutants that alter it and on phase shifts, and this focus has driven the identification of clock genes and clarified how clocks entrain to light--dark cycles. The waveform of the oscillation itself, by contrast, has attracted little attention as an indicator of the properties of the underlying oscillator. To assess the waveform directly, we focus on whether the trajectory in the plane spanned by a variable and its time derivative possesses an inflection point, and we define an oscillation to be harmonic when no inflection point is present. Bioluminescence recordings from cyanobacteria and from the mammalian SCN were harmonic in this sense, as were most of the core clock components in mathematical models of the circadian clock. Numerical analysis of the Goodwin model, a minimal representation of the core circadian oscillator, yielded harmonic oscillation throughout. We confirmed this numerical trend semi-analytically using a piecewise-linearized Goodwin model. Because it evaluates the properties of a waveform without assuming a model structure, the approach we propose offers a new perspective on the waveform analysis of biological rhythms in general, not only circadian ones.

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

25 extracted references · 16 canonical work pages

[11]

Sinusoidal regulation reduces circadian period variability. Sci. Reports 15, 33843. doi:10.1038/s41598-025-04614-z. Kim, J.K., Forger, D.B.,

doi:10.1038/s41598-025-04614-z
[16]

IEEE Trans

A theory of multiscale, curvature-based shape representation for planar curves. IEEE Trans. Pattern Anal. Mach. Intell. 14, 789–805. doi:10.1109/34.149591. Nobel Prize Outreach AB,

work page doi:10.1109/34.149591
[22]

volume 898 ofLecture Notes in Mathematics, pp

Springer, Berlin, Heidelberg. volume 898 ofLecture Notes in Mathematics, pp. 366–381. doi:10.1007/BFb0091924. Tanaka, Y., Fujita, Y., Macpherson, T., Harui, A., Kunisue, S., Setsu, G., Shao, X., Ono, D., Hikida, T., Fujita, Y., Matsuda, M., Okamura, H., Hasegawa, E., Doi, M.,

doi:10.1007/bfb0091924
[36]

Izhikevich, E.M.,

doi:10.1038/s44323-025-00054-0. Izhikevich, E.M.,

work page doi:10.1038/s44323-025-00054-0
[630]

13 Kondo, T.,

doi:10.1038/msb.2012.62. 13 Kondo, T.,

work page doi:10.1038/msb.2012.62 2012
[1967]

Biological rhythms and the behavior of populations of coupled oscillators. J. Theor. Biol. 16, 15–42. doi:10.1016/0022-5193(67)90051-3. Yoo, S.H., Yamazaki, S., Lowrey, P.L., Shimomura, K., Ko, C.H., Buhr, E.D., Siepka, S.M., Hong, H.K., Oh, W.J., Yoo, O.J., Menaker, M., Takahashi, J.S.,

work page doi:10.1016/0022-5193(67)90051-3
[1968]

negative feedback to one gene

Mathematics of cellular control processes i. negative feedback to one gene. J. Theor. Biol. 20, 202–208. doi:10.1016/0022-5193(68)90189-6. Imaizumi, T.,

work page doi:10.1016/0022-5193(68)90189-6
[1976]

the stability and lability of spontaneous frequency

A functional analysis of circadian pacemakers in nocturnal rodents: I. the stability and lability of spontaneous frequency. J. Comp. Physiol. 106, 223–252. doi:10.1007/BF01417856. de Ronde, W.H., Tostevin, F., ten Wolde, P.R.,

work page doi:10.1007/bf01417856

Show all 25 references

[1980]

Physical Review Letters 45, 712–716

Geometry from a time series. Physical Review Letters 45, 712–716. doi:10.1103/PhysRevLett.45.712. Pittendrigh, C.S., Daan, S.,

doi
[1992]

2nd ed., Cambridge University Press, Cambridge

Curves and Singularities: A Geometrical Introduction to Singularity Theory. 2nd ed., Cambridge University Press, Cambridge. doi:10.1017/CBO9781139172615. Bünning, E.,

doi
[1993]

Circadian rhythms in prokaryotes: Luciferase as a reporter of circadian gene expression in cyanobacteria. Proc. Natl. Acad. Sci. United States of America 90, 5672–5676. doi:10.1073/pnas.90.12.5672. Kuramoto, Y.,

doi
[1995]

Proceedings of the Royal Society B: Biological Sciences 261, 319–324

A model for circadian oscillations in the Drosophila period protein (PER). Proceedings of the Royal Society B: Biological Sciences 261, 319–324. doi:10.1098/rspb.1995.0153. Goodwin, B.C.,

1995
[2002]

Comparative study of circadian clock models, in search of processes promoting oscillation. J. Theor. Biol. 216, 193–208. doi:10.1006/jtbi.2002.2546. Mokhtarian, F., Mackworth, A.K.,

2002
[2004]

PERIOD2::LUCIFERASE real-time reporting of circadian dynamics reveals persistent circadian oscillations in mouse peripheral tissues. Proc. Natl. Acad. Sci. United States of America 101, 5339–5346. doi:10.1073/pnas.0308709101. 14 Fig. 1 5 −3 6 −10 3 −2 3 −3 5 −3 Stuart–Landau F...

doi
[2007]

MIT Press, Cambridge, MA

Dynamical Systems in Neuroscience. MIT Press, Cambridge, MA. doi:10.7551/mitpress/ 2526.001.0001. Johnson, C.H., Elliott, J., Foster, R., Honma, K., Kronauer, R.,

doi
[2009]

volume 66 ofWorld Scientific Series on Nonlinear Science, Series A

Differential Geometry Applied to Dynamical Systems. volume 66 ofWorld Scientific Series on Nonlinear Science, Series A. World Scientific, Singapore. doi:10.1142/7333. Ginoux, J.M., Rossetto, B., Chua, L.O.,

doi
[2010]

Effect of feedback on the fidelity of information transmission of time-varying signals. Phys. Rev. E 82, 031914. doi:10.1103/PhysRevE.82.031914. Sasai, M.,

doi
[2011]

ISRN Appl

Numerical differentiation of noisy, nonsmooth data. ISRN Appl. Math. 2011, 164564. doi:10. 5402/2011/164564. Gibo, S., Kunihiro, T., Hatsuda, T., Kurosawa, G.,

2011
[2017]

Accessed: 2 August

Discoveries of molecular mechanisms controlling the circadian rhythm.https: //www.nobelprize.org/prizes/medicine/2017/advanced-information/. Accessed: 2 August

2017
[2018]

Cell Metabolism 27, 854–868.e8

A transcriptional circuit filters oscillating circadian hormonal inputs to regulate fat cell differentiation. Cell Metabolism 27, 854–868.e8. doi:10.1016/j.cmet.2018.03.012. Bruce, J.W., Giblin, P.J.,

2018 doi
[2019]

Non-sinusoidal waveform in temperature-compensated circadian oscillations. Biophys. J. 116, 741–751. doi:10.1016/j.bpj.2018.12.022. Ginoux, J.M.,

2018 doi
[2021]

(Eds.), Circadian Rhythms in Bacteria and Microbiomes, pp

Around the circadian clock: Review and preview, in: Johnson, C.H., Rust, M.J. (Eds.), Circadian Rhythms in Bacteria and Microbiomes, pp. 21–52. doi:10.1007/978-3-030-72158-9_2. Kondo, T., Strayer, C.A., Kulkarni, R.D., Taylor, W., Ishiura, M., Golden, S.S., Johnson, C.H.,

doi
[2022]

PLOS Comput

Role of the reaction-structure coupling in temperature compensation of the KaiABC circadian rhythm. PLOS Comput. Biol. 18, e1010494. doi:10.1371/journal.pcbi.1010494. Takens, F.,

doi
[2025]

PLOS Comput

Waveform distortion for temperature compensation and synchronization in circadian rhythms: An approach based on the renormalization group method. PLOS Comput. Biol. 21, e1013246. doi:10.1371/journal.pcbi.1013246. Gibo, S., Kurosawa, G.,

doi
[2026]

doi:10.64898/2026.05.23.727383

The slow-acting G-protein Gz defines the duration of circadian rest time. doi:10.64898/2026.05.23.727383. bioRxiv preprint. Winfree, A.T.,

2026 doi

Pith tools

Reviewed August 4, 2026 · model on record in the stance chip above.

[1] [11]

Sinusoidal regulation reduces circadian period variability. Sci. Reports 15, 33843. doi:10.1038/s41598-025-04614-z. Kim, J.K., Forger, D.B.,

doi:10.1038/s41598-025-04614-z

[2] [16]

IEEE Trans

A theory of multiscale, curvature-based shape representation for planar curves. IEEE Trans. Pattern Anal. Mach. Intell. 14, 789–805. doi:10.1109/34.149591. Nobel Prize Outreach AB,

work page doi:10.1109/34.149591

[3] [22]

volume 898 ofLecture Notes in Mathematics, pp

Springer, Berlin, Heidelberg. volume 898 ofLecture Notes in Mathematics, pp. 366–381. doi:10.1007/BFb0091924. Tanaka, Y., Fujita, Y., Macpherson, T., Harui, A., Kunisue, S., Setsu, G., Shao, X., Ono, D., Hikida, T., Fujita, Y., Matsuda, M., Okamura, H., Hasegawa, E., Doi, M.,

doi:10.1007/bfb0091924

[4] [36]

Izhikevich, E.M.,

doi:10.1038/s44323-025-00054-0. Izhikevich, E.M.,

work page doi:10.1038/s44323-025-00054-0

[5] [630]

13 Kondo, T.,

doi:10.1038/msb.2012.62. 13 Kondo, T.,

work page doi:10.1038/msb.2012.62 2012

[6] [1967]

Biological rhythms and the behavior of populations of coupled oscillators. J. Theor. Biol. 16, 15–42. doi:10.1016/0022-5193(67)90051-3. Yoo, S.H., Yamazaki, S., Lowrey, P.L., Shimomura, K., Ko, C.H., Buhr, E.D., Siepka, S.M., Hong, H.K., Oh, W.J., Yoo, O.J., Menaker, M., Takahashi, J.S.,

work page doi:10.1016/0022-5193(67)90051-3

[7] [1968]

negative feedback to one gene

Mathematics of cellular control processes i. negative feedback to one gene. J. Theor. Biol. 20, 202–208. doi:10.1016/0022-5193(68)90189-6. Imaizumi, T.,

work page doi:10.1016/0022-5193(68)90189-6

[8] [1976]

the stability and lability of spontaneous frequency

A functional analysis of circadian pacemakers in nocturnal rodents: I. the stability and lability of spontaneous frequency. J. Comp. Physiol. 106, 223–252. doi:10.1007/BF01417856. de Ronde, W.H., Tostevin, F., ten Wolde, P.R.,

work page doi:10.1007/bf01417856