pith. sign in

arxiv: 2406.03456 · v2 · submitted 2024-06-05 · 🧬 q-bio.MN · math.DS

Recurrent neural chemical reaction networks that approximate arbitrary dynamics

Alexander Dack , Benjamin Qureshi , Thomas E. Ouldridge , Tomislav Plesa This is my paper

Pith reviewed 2026-05-24 00:26 UTC · model grok-4.3

classification 🧬 q-bio.MN math.DS
keywords chemical reaction networksrecurrent neural networkssynthetic biologyDNA strand displacementdynamical systemsoscillationsmulti-stabilitychaos
0
0 comments X

The pith

A modular network of chemical neurons can be trained to produce any desired dynamics.

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

The paper puts forward recurrent neural chemical reaction networks, built from modular chemical neurons, as a way to construct synthetic chemical systems that exhibit complex behaviors like multi-stability, oscillations, and chaos. It proves that enough neurons combined with sufficiently fast reactions allow systematic training to match any target dynamics. This approach matters because designing such rich dynamics from scratch has been a central obstacle in synthetic biology. The work further shows that modest-sized networks with practical reaction rates can be trained for key biological features and realized using DNA strand displacement.

Core claim

With sufficiently many chemical neurons and suitably fast reactions, the RNCRN can be systematically trained to achieve any dynamics. RNCRNs with relatively small numbers of chemical neurons and moderate reaction rates are trained to display biologically important dynamical features, and the architecture is shown to be experimentally implementable with DNA-strand-displacement technologies.

What carries the argument

The modular network of chemical neurons, each performing signal processing through mass-action chemical reactions arranged to emulate recurrent neural network computations.

If this is right

  • RNCRNs can generate multi-stability, oscillations, and chaos on demand.
  • Small RNCRNs with practical reaction rates suffice for many biologically relevant behaviors.
  • DNA strand displacement provides a concrete experimental route to implement the trained networks.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The same training procedure could be applied to design chemical controllers that respond to changing inputs in a programmable way.
  • Connections between neural network training algorithms and chemical rate tuning might allow automated design tools for molecular circuits.
  • If the modularity holds, the method could extend to hybrid systems that combine chemical computation with other molecular components.

Load-bearing premise

The modular chemical neurons can be physically realized and combined so their reaction rates remain independent and tunable without cross-talk or side reactions altering the mass-action dynamics.

What would settle it

An attempt to build and run a small trained RNCRN via DNA strand displacement that fails to produce the intended dynamics due to unintended reactions or rate interference.

Figures

Figures reproduced from arXiv: 2406.03456 by Alexander Dack, Benjamin Qureshi, Thomas E. Ouldridge, Tomislav Plesa.

Figure 1
Figure 1. Figure 1: A visualization of an RNCRN with executive species [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: RNCRN approximations of the multi-stable target system (8). (a) The vector field of [PITH_FULL_IMAGE:figures/full_fig_p008_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: DNA-strand-displacement implementation of the RNCRN with RREs (10) with [PITH_FULL_IMAGE:figures/full_fig_p010_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: RNCRN approximation of the oscillatory target system (12). (a) The vector field of (12) [PITH_FULL_IMAGE:figures/full_fig_p011_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Robustness of RNCRNs approximating oscillatory target system (12). (a) Proportion of [PITH_FULL_IMAGE:figures/full_fig_p012_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: RNCRN approximation of the target system (13) with a chaotic attractor. (a) Solution [PITH_FULL_IMAGE:figures/full_fig_p013_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: RNCRN approximations of the multi-stable target system (8). (a)-(e) The vector field of the [PITH_FULL_IMAGE:figures/full_fig_p026_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: DNA-strand-displacement implementation of the RNCRN with RREs (10) with [PITH_FULL_IMAGE:figures/full_fig_p029_8.png] view at source ↗
Figure 9
Figure 9. Figure 9: RNCRN approximation of the oscillatory target system (12) trained with noisy Algorithm 1. [PITH_FULL_IMAGE:figures/full_fig_p033_9.png] view at source ↗
Figure 10
Figure 10. Figure 10: Initial-condition sensitivity of system (49) that approximates target system (8). Panel (a) [PITH_FULL_IMAGE:figures/full_fig_p036_10.png] view at source ↗
read the original abstract

Many important phenomena in biochemistry and biology exploit dynamical features such as multi-stability, oscillations, and chaos. Construction of novel chemical systems with such rich dynamics is a challenging problem central to the fields of synthetic biology and molecular nanotechnology. In this paper, we address this problem by putting forward a molecular version of a recurrent artificial neural network, which we call recurrent neural chemical reaction network (RNCRN). The RNCRN uses a modular architecture - a network of chemical neurons - to approximate arbitrary dynamics. We first prove that with sufficiently many chemical neurons and suitably fast reactions, the RNCRN can be systematically trained to achieve any dynamics. RNCRNs with relatively small number of chemical neurons and a moderate range of reaction rates are then trained to display a variety of biologically-important dynamical features. We also demonstrate that such RNCRNs are experimentally implementable with DNA-strand-displacement technologies.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

1 major / 1 minor

Summary. The manuscript introduces recurrent neural chemical reaction networks (RNCRNs), a modular architecture of chemical neurons governed by mass-action ODEs. It proves that sufficiently many neurons with suitably fast reactions can be trained to approximate arbitrary dynamics, demonstrates successful training of small networks to produce multi-stability, oscillations, and chaos, and provides a DNA strand-displacement construction claimed to realize the networks experimentally.

Significance. If the universal-approximation result and the DNA implementation both hold, the work supplies a systematic, trainable route to engineering chemical systems with prescribed complex dynamics. The explicit modular construction and the proof of approximation (when the separation-of-timescales assumption is satisfied) are concrete strengths that could be leveraged in synthetic biology.

major comments (1)
  1. [Implementation section (DNA-strand-displacement construction)] Implementation section (DNA-strand-displacement construction): the claim that concatenated networks preserve exactly the mass-action kinetics of the isolated modular neurons rests on the unverified assumption that no additional reactions or rate modifications arise upon concatenation. The supplied candidate circuit does not include an explicit check (e.g., enumeration of all possible strand interactions or effective-rate derivation) that would confirm the effective ODEs remain identical to the mathematical RNCRN model; this assumption is load-bearing for transferring the approximation theorem to a physical realization.
minor comments (1)
  1. Notation for reaction rates and neuron modules should be unified across the mathematical model and the DNA construction to avoid ambiguity when mapping parameters.

Simulated Author's Rebuttal

1 responses · 0 unresolved

We thank the referee for their careful reading of the manuscript and for highlighting an important point about the DNA implementation. We address the major comment below.

read point-by-point responses

  1. Referee: Implementation section (DNA-strand-displacement construction): the claim that concatenated networks preserve exactly the mass-action kinetics of the isolated modular neurons rests on the unverified assumption that no additional reactions or rate modifications arise upon concatenation. The supplied candidate circuit does not include an explicit check (e.g., enumeration of all possible strand interactions or effective-rate derivation) that would confirm the effective ODEs remain identical to the mathematical RNCRN model; this assumption is load-bearing for transferring the approximation theorem to a physical realization.

    Authors: We agree that the manuscript presents the DNA-strand-displacement construction as a candidate circuit without performing an exhaustive enumeration of all possible strand interactions or deriving effective rates for the concatenated system. The design relies on standard practices in the field (orthogonal domains, unique toehold sequences, and leakless reaction motifs drawn from prior DNA computing literature) to ensure that modules do not introduce new reactions or alter rates. However, these practices are not accompanied by an explicit verification step in the current text. In the revised manuscript we will add a dedicated paragraph in the implementation section that (i) states the modularity assumption explicitly, (ii) cites the sequence-design heuristics used to minimize crosstalk, and (iii) notes that a full interaction enumeration is feasible for the small networks considered but was omitted to maintain focus on the approximation theorem and training results. We will also indicate that such a check can be performed with existing tools (e.g., Visual DSD or similar) prior to experimental realization. revision: partial

Circularity Check

0 steps flagged

No significant circularity in the RNCRN universal approximation proof

full rationale

The paper's central claim rests on a stated mathematical proof that sufficiently many chemical neurons with fast reactions allow systematic training to achieve arbitrary dynamics. This proof operates on the defined mass-action ODEs of the modular architecture and does not reduce by construction to any fitted parameter, self-definition, or self-citation chain. The DNA strand-displacement implementation is presented as a separate demonstration of realizability rather than a load-bearing step in the approximation theorem. The derivation is therefore self-contained and independent of the paper's own inputs.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The framework assumes that chemical reaction networks can be engineered with independent, arbitrarily tunable rates and that timescale separation between fast internal reactions and slower target dynamics can be maintained.

axioms (1)
  • domain assumption Chemical reactions can be designed and combined modularly with independent, arbitrarily assignable rates without cross-talk.
    Required for both the universal approximation theorem and the DNA implementation claim.

pith-pipeline@v0.9.0 · 5689 in / 1113 out tokens · 20906 ms · 2026-05-24T00:26:37.117848+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

What do these tags mean?
matches
The paper's claim is directly supported by a theorem in the formal canon.
supports
The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends
The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses
The paper appears to rely on the theorem as machinery.
contradicts
The paper's claim conflicts with a theorem or certificate in the canon.
unclear
Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Reference graph

Works this paper leans on

87 extracted references · 87 canonical work pages

  1. [1]

    A logical calculus of ideas immanent in nervous activity,

    W. S. McCulloch and W. Pitts, “A logical calculus of ideas immanent in nervous activity,” Bulletin of Mathematical Biophysics, vol. 5, pp. 115–133, 1943

    work page 1943

  2. [2]

    Approximation by superpositions of a sigmoidal function,

    G. Cybenko, “Approximation by superpositions of a sigmoidal function,”Mathematics of Control, Signals, and Systems, vol. 2, pp. 303–314, 1989

    work page 1989

  3. [3]

    Multilayer feedforward networks are universal approximators,

    K. Hornik, M. Stinchcombe, and H. White, “Multilayer feedforward networks are universal approximators,”Neural Networks, vol. 2, pp. 359–366, 1989

    work page 1989

  4. [4]

    Approximation theory of the MLP model in neural networks,

    A. Pinkus, “Approximation theory of the MLP model in neural networks,”Acta Numerica, vol. 8, pp. 143–195, 1999

    work page 1999

  5. [5]

    Deep learning,

    Y. LeCun, Y. Bengio, and G. Hinton, “Deep learning,”Nature, vol. 521, pp. 436–444, 2015

    work page 2015

  6. [6]

    Neural Ordinary Differential Equations,

    R. T. Q. Chen, Y. Rubanova, J. Bettencourt, and D. Duvenaud, “Neural Ordinary Differential Equations,” 2018

    work page 2018

  7. [7]

    Neurons with graded response have collective computational properties like those of two-state neurons.,

    J. J. Hopfield, “Neurons with graded response have collective computational properties like those of two-state neurons.,”Proceedings of the National Academy of Sciences, vol. 81, pp. 3088–3092, May 1984

    work page 1984

  8. [8]

    Large Associative Memory Problem in Neurobiology and Machine Learning,

    D. Krotov and J. Hopfield, “Large Associative Memory Problem in Neurobiology and Machine Learning,” Apr. 2021. arXiv:2008.06996

    work page arXiv 2021

  9. [9]

    Differential equations and dynamical systems,

    L. Perko, “Differential equations and dynamical systems,”3rd Edition, Springer-Verlag, 2001

    work page 2001

  10. [10]

    Approximation of dynamical systems by continuous time recurrent neural networks,

    K.-i. Funahashi and Y. Nakamura, “Approximation of dynamical systems by continuous time recurrent neural networks,”Neural Networks, vol. 6, pp. 801–806, Jan. 1993

    work page 1993

  11. [11]

    Model-Free Prediction of Large Spatiotempo- rally Chaotic Systems from Data: A Reservoir Computing Approach,

    J. Pathak, B. Hunt, M. Girvan, Z. Lu, and E. Ott, “Model-Free Prediction of Large Spatiotempo- rally Chaotic Systems from Data: A Reservoir Computing Approach,”Physical Review Letters, vol. 120, p. 024102, Jan. 2018

    work page 2018

  12. [12]

    Learning dynamical systems by recurrent neural networks from orbits,

    M. Kimura and R. Nakano, “Learning dynamical systems by recurrent neural networks from orbits,”Neural Networks, vol. 11, pp. 1589–1599, Dec. 1998

    work page 1998

  13. [13]

    Learning nonlinear dynamics by recurrent neural networks (some problems on the theory of dynamical systems in applied sciences),

    M.-a. Sato and Y. Murakami, “Learning nonlinear dynamics by recurrent neural networks (some problems on the theory of dynamical systems in applied sciences),”数理解析研究所講究録, vol. 760, pp. 71–87, 1991

    work page 1991

  14. [14]

    Con- trol of chaotic systems by deep reinforcement learning,

    M. A. Bucci, O. Semeraro, A. Allauzen, G. Wisniewski, L. Cordier, and L. Mathelin, “Con- trol of chaotic systems by deep reinforcement learning,”Proceedings of the Royal Society A: Mathematical, Physical and Engineering Sciences, vol. 475, p. 20190351, 2019

    work page 2019

  15. [15]

    From electric circuits to chemical networks,

    L. Cardelli, M. Tribastone, and M. Tschaikowski, “From electric circuits to chemical networks,” Natural Computing, vol. 19, pp. 237–248, 2020

    work page 2020

  16. [16]

    Computational functions in biochemical reaction networks,

    A. Arkin and J. Ross, “Computational functions in biochemical reaction networks,”Biophysical Journal, vol. 67, pp. 560–578, 1994

    work page 1994

  17. [17]

    Enzyme-Free Nucleic Acid Logic Circuits,

    G. Seelig, D. Soloveichik, D. Y. Zhang, and E. Winfree, “Enzyme-Free Nucleic Acid Logic Circuits,”Science, vol. 314, pp. 1585–1588, 2006. 16

    work page 2006

  18. [18]

    ScalingUpDigitalCircuitComputationwithDNAStrandDisplacement Cascades,

    L.QianandE.Winfree, “ScalingUpDigitalCircuitComputationwithDNAStrandDisplacement Cascades,”Science, vol. 332, pp. 1196–1201, 2011

    work page 2011

  19. [19]

    Control theory meets synthetic biology,

    D. Del Vecchio, A. J. Dy, and Y. Qian, “Control theory meets synthetic biology,”Journal of The Royal Society Interface, vol. 13, p. 20160380, 2016

    work page 2016

  20. [20]

    Antithetic Integral Feedback Ensures Robust Perfect Adaptation in Noisy Biomolecular Networks,

    C. Briat, A. Gupta, and M. Khammash, “Antithetic Integral Feedback Ensures Robust Perfect Adaptation in Noisy Biomolecular Networks,”Cell Systems, vol. 2, pp. 15–26, 2016

    work page 2016

  21. [21]

    Quasi-robust control of biochemical reaction networks via stochastic morphing,

    T. Plesa, G.-B. Stan, T. E. Ouldridge, and W. Bae, “Quasi-robust control of biochemical reaction networks via stochastic morphing,”Journal of The Royal Society Interface, vol. 18, p. 20200985, 2021

    work page 2021

  22. [22]

    Lecture notes on chemical reaction networks,

    M. Feinberg, “Lecture notes on chemical reaction networks,”Lecture Notes, Mathematics Research Center, University of Wisconsin-Madison, 1979

    work page 1979

  23. [23]

    Chemical implementation of neural networks and Turing machines.,

    A. Hjelmfelt, E. D. Weinberger, and J. Ross, “Chemical implementation of neural networks and Turing machines.,”Proceedings of the National Academy of Sciences, vol. 88, pp. 10983–10987, 1991

    work page 1991

  24. [24]

    Neural network computation with DNA strand displacement cascades,

    L. Qian, E. Winfree, and J. Bruck, “Neural network computation with DNA strand displacement cascades,”Nature, vol. 475, pp. 368–372, 2011

    work page 2011

  25. [25]

    Neural Network Computation by In Vitro Transcriptional Circuits,

    J. Kim, J. Hopfield, and E. Winfree, “Neural Network Computation by In Vitro Transcriptional Circuits,”Advances in Neural Information Processing Systems, vol. 17, 2004

    work page 2004

  26. [26]

    Reconfigurable neuromorphic computation in biochemical systems,

    H.-J. K. Chiang, J.-H. R. Jiang, and F. Fages, “Reconfigurable neuromorphic computation in biochemical systems,” in2015 37th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC), pp. 937–940, 2015

    work page 2015

  27. [27]

    Programming and training rate-independent chemical reaction networks,

    M. Vasić, C. Chalk, A. Luchsinger, S. Khurshid, and D. Soloveichik, “Programming and training rate-independent chemical reaction networks,”Proceedings of the National Academy of Sciences, vol. 119, p. e2111552119, 2022

    work page 2022

  28. [28]

    Robust Digital Molecular Design of Binarized Neural Networks,

    J. Linder, Y.-J. Chen, D. Wong, G. Seelig, L. Ceze, and K. Strauss, “Robust Digital Molecular Design of Binarized Neural Networks,” in27th International Conference on DNA Computing and Molecular Programming (DNA 27), vol. 205, pp. 1:1–1:20, 2021

    work page 2021

  29. [29]

    On reaction network implementations of neural networks,

    D. F. Anderson, B. Joshi, and A. Deshpande, “On reaction network implementations of neural networks,”Journal of The Royal Society Interface, vol. 18, p. 20210031, 2021

    work page 2021

  30. [30]

    NeuralCRNs: A Natural Implementation of Learning in Chemical Reaction Networks,

    R. T. Nagipogu and J. H. Reif, “NeuralCRNs: A Natural Implementation of Learning in Chemical Reaction Networks,” Aug. 2024. arXiv:2409.00034

    work page arXiv 2024

  31. [31]

    Molecular Computation for Molecular Classification,

    C. Kieffer, A. J. Genot, Y. Rondelez, and G. Gines, “Molecular Computation for Molecular Classification,”Advanced Biology, vol. 7, p. 2200203, 2023

    work page 2023

  32. [32]

    Neural networks built from enzymatic reactions can operate as linear and nonlinear classifiers,

    C. C. Samaniego, E. Wallace, F. Blanchini, E. Franco, and G. Giordano, “Neural networks built from enzymatic reactions can operate as linear and nonlinear classifiers,”bioRxiv, 2024

    work page 2024

  33. [33]

    DNA Input Classification by a Riboregulator-Based Cell-Free Perceptron,

    A. J. Van Der Linden, P. A. Pieters, M. W. Bartelds, B. L. Nathalia, P. Yin, W. T. S. Huck, J. Kim, and T. F. A. De Greef, “DNA Input Classification by a Riboregulator-Based Cell-Free Perceptron,”ACS Synthetic Biology, vol. 11, pp. 1510–1520, 2022. 17

    work page 2022

  34. [34]

    Chemical Boltzmann Machines,

    W. Poole, A. Ortiz-Muñoz, A. Behera, N. S. Jones, T. E. Ouldridge, E. Winfree, and M. Gopalkr- ishnan, “Chemical Boltzmann Machines,” inDNA 2017: DNA Computing and Molecular Programming, vol. 10467, pp. 210–231, 2017

    work page 2017

  35. [35]

    Scaling up molecular pattern recognition with DNA-based winner- take-all neural networks,

    K. M. Cherry and L. Qian, “Scaling up molecular pattern recognition with DNA-based winner- take-all neural networks,”Nature, vol. 559, pp. 370–376, 2018

    work page 2018

  36. [36]

    Molecular convolutional neural networks with DNA regulatory circuits,

    X. Xiong, T. Zhu, Y. Zhu, M. Cao, J. Xiao, L. Li, F. Wang, C. Fan, and H. Pei, “Molecular convolutional neural networks with DNA regulatory circuits,”Nature Machine Intelligence, vol. 4, pp. 625–635, 2022

    work page 2022

  37. [37]

    A Dynamical Biomolecular Neural Network,

    A. Moorman, C. C. Samaniego, C. Maley, and R. Weiss, “A Dynamical Biomolecular Neural Network,” in2019 IEEE 58th Conference on Decision and Control (CDC), pp. 1797–1802, 2019. ISSN: 2576-2370

    work page 2019

  38. [38]

    A molecular multi-gene classifier for disease diagnostics,

    R. Lopez, R. Wang, and G. Seelig, “A molecular multi-gene classifier for disease diagnostics,” Nature Chemistry, vol. 10, pp. 746–754, 2018

    work page 2018

  39. [39]

    Scaling down DNA circuits with competitive neural networks,

    A. J. Genot, T. Fujii, and Y. Rondelez, “Scaling down DNA circuits with competitive neural networks,”Journal of The Royal Society Interface, vol. 10, p. 20130212, 2013

    work page 2013

  40. [40]

    Metabolic perceptrons for neural computing in biological systems,

    A. Pandi, M. Koch, P. L. Voyvodic, P. Soudier, J. Bonnet, M. Kushwaha, and J.-L. Faulon, “Metabolic perceptrons for neural computing in biological systems,”Nature Communications, vol. 10, p. 3880, 2019

    work page 2019

  41. [41]

    Synthetic neural-like computing in microbial consortia for pattern recognition,

    X. Li, L. Rizik, V. Kravchik, M. Khoury, N. Korin, and R. Daniel, “Synthetic neural-like computing in microbial consortia for pattern recognition,”Nature Communications, vol. 12, p. 3139, 2021

    work page 2021

  42. [42]

    Nonlinear decision-making with enzymatic neural networks,

    S. Okumura, G. Gines, N. Lobato-Dauzier, A. Baccouche, R. Deteix, T. Fujii, Y. Rondelez, and A. J. Genot, “Nonlinear decision-making with enzymatic neural networks,”Nature, vol. 610, pp. 496–501, 2022

    work page 2022

  43. [43]

    Rosenblatt,The perceptron, a perceiving and recognizing automaton Project Para

    F. Rosenblatt,The perceptron, a perceiving and recognizing automaton Project Para. Cornell Aeronautical Laboratory, 1957

    work page 1957

  44. [44]

    Information Processing in Biochemical Networks,

    G. Tkačik and P. R. T. Wolde, “Information Processing in Biochemical Networks,”Annual Review of Biophysics, Feb. 2025

    work page 2025

  45. [45]

    A positive-feedback-based bistable ‘memory module’ that governs a cell fate decision,

    W. Xiong and J. E. Ferrell, “A positive-feedback-based bistable ‘memory module’ that governs a cell fate decision,”Nature, vol. 426, pp. 460–465, Nov. 2003

    work page 2003

  46. [46]

    Feedback of the Drosophila period gene product on circadian cycling of its messenger RNA levels,

    P. E. Hardin, J. C. Hall, and M. Rosbash, “Feedback of the Drosophila period gene product on circadian cycling of its messenger RNA levels,”Nature, vol. 343, pp. 536–540, Feb. 1990

    work page 1990

  47. [47]

    Generation of oscillations by the p53-Mdm2 feedback loop: A theoretical and experimental study,

    R. Lev Bar-Or, R. Maya, L. A. Segel, U. Alon, A. J. Levine, and M. Oren, “Generation of oscillations by the p53-Mdm2 feedback loop: A theoretical and experimental study,”Proceedings of the National Academy of Sciences, vol. 97, pp. 11250–11255, Oct. 2000

    work page 2000

  48. [48]

    On chaotic dynamics in transcription factors and the associated effects in differential gene regulation,

    M. L. Heltberg, S. Krishna, and M. H. Jensen, “On chaotic dynamics in transcription factors and the associated effects in differential gene regulation,”Nature Communications, vol. 10, p. 71, Jan. 2019. 18

    work page 2019

  49. [49]

    Chaos in a bacterial stress response,

    D. Choudhary, K. R. Foster, and S. Uphoff, “Chaos in a bacterial stress response,”Current Biology, vol. 33, pp. 5404–5414.e9, Dec. 2023

    work page 2023

  50. [50]

    Enzyme-free nucleic acid dynamical systems,

    N. Srinivas, J. Parkin, G. Seelig, E. Winfree, and D. Soloveichik, “Enzyme-free nucleic acid dynamical systems,”Science, vol. 358, p. eaal2052, Dec. 2017

    work page 2017

  51. [51]

    DNA Strand-Displacement Temporal Logic Circuits,

    A. P. Lapteva, N. Sarraf, and L. Qian, “DNA Strand-Displacement Temporal Logic Circuits,” Journal of the American Chemical Society, vol. 144, pp. 12443–12449, July 2022

    work page 2022

  52. [52]

    Predator–Prey Molecular Ecosystems,

    T. Fujii and Y. Rondelez, “Predator–Prey Molecular Ecosystems,”ACS Nano, vol. 7, pp. 27–34, Jan. 2013

    work page 2013

  53. [53]

    Programmable autonomous synthesis of single-stranded DNA,

    J. Y. Kishi, T. E. Schaus, N. Gopalkrishnan, F. Xuan, and P. Yin, “Programmable autonomous synthesis of single-stranded DNA,”Nature Chemistry, vol. 10, pp. 155–164, Feb. 2018

    work page 2018

  54. [54]

    Neural coding of temperature with a DNA-based spiking chemical neuron,

    N. Lobato-Dauzier, A. Baccouche, G. Gines, T. Levi, Y. Rondelez, T. Fujii, S. H. Kim, N. Aubert- Kato, and A. J. Genot, “Neural coding of temperature with a DNA-based spiking chemical neuron,”Nature Chemical Engineering, vol. 1, pp. 510–521, Aug. 2024

    work page 2024

  55. [55]

    A fast, robust and tunable synthetic gene oscillator,

    J. Stricker, S. Cookson, M. R. Bennett, W. H. Mather, L. S. Tsimring, and J. Hasty, “A fast, robust and tunable synthetic gene oscillator,”Nature, vol. 456, pp. 516–519, Nov. 2008

    work page 2008

  56. [56]

    Synthetic multistability in mammalian cells,

    R. Zhu, J. M. Del Rio-Salgado, J. Garcia-Ojalvo, and M. B. Elowitz, “Synthetic multistability in mammalian cells,”Science, vol. 375, p. eabg9765, Jan. 2022

    work page 2022

  57. [57]

    DNA Reservoir Computing: A Novel Molecular Computing Approach,

    A. Goudarzi, M. R. Lakin, and D. Stefanovic, “DNA Reservoir Computing: A Novel Molecular Computing Approach,” inDNA Computing and Molecular Programming(D. Soloveichik and B. Yurke, eds.), vol. 8141, pp. 76–89, Cham: Springer International Publishing, 2013

    work page 2013

  58. [58]

    Chemical reservoir computation in a self-organizing reaction network,

    M. G. Baltussen, T. J. De Jong, Q. Duez, W. E. Robinson, and W. T. S. Huck, “Chemical reservoir computation in a self-organizing reaction network,”Nature, vol. 631, pp. 549–555, July 2024

    work page 2024

  59. [59]

    Chemical reaction systems with a homoclinic bifurcation: an inverse problem,

    T. Plesa, T. Vejchodský, and R. Erban, “Chemical reaction systems with a homoclinic bifurcation: an inverse problem,”Journal of Mathematical Chemistry, vol. 54, pp. 1884–1915, 2016

    work page 1915

  60. [60]

    Chemical systems consisting only of elementary steps – a paradigma for nonlinear behavior,

    T. Wilhelm, “Chemical systems consisting only of elementary steps – a paradigma for nonlinear behavior,”Journal of Mathematical Chemistry, vol. 27, pp. 71–88, 2000

    work page 2000

  61. [61]

    Stochastic approximations of higher-molecular by bi-molecular reactions,

    T. Plesa, “Stochastic approximations of higher-molecular by bi-molecular reactions,”Journal of Mathematical Biology, vol. 86, p. 28, 2023

    work page 2023

  62. [62]

    DNA as a universal substrate for chemical kinetics,

    D. Soloveichik, G. Seelig, and E. Winfree, “DNA as a universal substrate for chemical kinetics,” Proceedings of the National Academy of Sciences, vol. 107, pp. 5393–5398, 2010

    work page 2010

  63. [63]

    Programmable chemical controllers made from DNA,

    Y.-J. Chen, N. Dalchau, N. Srinivas, A. Phillips, L. Cardelli, D. Soloveichik, and G. Seelig, “Programmable chemical controllers made from DNA,”Nature Nanotechnology, vol. 8, pp. 755– 762, Oct. 2013

    work page 2013

  64. [64]

    Implementing Arbitrary CRNs Using Strand Displacing Polymerase,

    S. Shah, T. Song, X. Song, M. Yang, and J. Reif, “Implementing Arbitrary CRNs Using Strand Displacing Polymerase,” inDNA Computing and Molecular Programming(C. Thachuk and Y. Liu, eds.), vol. 11648, pp. 21–36, Cham: Springer International Publishing, 2019. 19

    work page 2019

  65. [65]

    Learning representations by back- propagating errors,

    D. E. Rumelhart, G. E. Hinton, and R. J. Williams, “Learning representations by back- propagating errors,”Nature, vol. 323, pp. 533–536, Oct. 1986

    work page 1986

  66. [66]

    Chaotic oscillations and cycles in multi-trophic ecological systems,

    L. Stone and D. He, “Chaotic oscillations and cycles in multi-trophic ecological systems,”Journal of Theoretical Biology, vol. 248, pp. 382–390, 2007

    work page 2007

  67. [67]

    Mapping dynamical systems into chemical reactions,

    T. Plesa, “Mapping dynamical systems into chemical reactions,” Available as https://arxiv.org/pdf/2406.03473, 2025

    work page arXiv 2025

  68. [68]

    Nonlinear chemical kinetic schemes derived from mechanical and electrical dynamical systems,

    N. Samardzija, L. D. Greller, and E. Wasserman, “Nonlinear chemical kinetic schemes derived from mechanical and electrical dynamical systems,”The Journal of Chemical Physics, vol. 90, pp. 2296–2304, 1989

    work page 1989

  69. [69]

    Cooperativecatalysisandchemicalchaos: achemicalmodelfortheLorenzequations,

    D.Poland, “Cooperativecatalysisandchemicalchaos: achemicalmodelfortheLorenzequations,” Physica D: Nonlinear Phenomena, vol. 65, pp. 86–99, 1993

    work page 1993

  70. [70]

    Test Models for Statistical Inference: Two-Dimensional Reaction Systems Displaying Limit Cycle Bifurcations and Bistability,

    T. Plesa, T. Vejchodský, and R. Erban, “Test Models for Statistical Inference: Two-Dimensional Reaction Systems Displaying Limit Cycle Bifurcations and Bistability,” inStochastic Processes, Multiscale Modeling, and Numerical Methods for Computational Cellular Biology, pp. 3–27, 2017

    work page 2017

  71. [71]

    Integral feedback in synthetic biology: Negative- equilibrium catastrophe (Appendix D),

    T. Plesa, A. Dack, and T. E. Ouldridge, “Integral feedback in synthetic biology: Negative- equilibrium catastrophe (Appendix D),”Journal of Mathematical Chemistry, vol. 61, pp. 1980– 2018, 2023

    work page 1980

  72. [72]

    Mass action realizations of reaction kinetic system models on various time scales,

    K. M. Hangos and G. Szederkényi, “Mass action realizations of reaction kinetic system models on various time scales,”Journal of Physics: Conference Series, vol. 268, p. 012009, 2011

    work page 2011

  73. [73]

    Universal formats for nonlinear dynamical systems,

    K. Kowalski, “Universal formats for nonlinear dynamical systems,”Chemical Physics Letters, vol. 209, pp. 167–170, 1993

    work page 1993

  74. [74]

    Universal formats for nonlinear ordinary differential systems,

    E. H. Kerner, “Universal formats for nonlinear ordinary differential systems,”Journal of Mathematical Physics, vol. 22, pp. 1366–1371, 1981

    work page 1981

  75. [75]

    Control of dna strand displacement kinetics using toehold exchange,

    D. Y. Zhang and E. Winfree, “Control of dna strand displacement kinetics using toehold exchange,”Journal of the American Chemical Society, vol. 131, no. 47, pp. 17303–17314, 2009. PMID: 19894722

    work page 2009

  76. [76]

    Design of hidden thermodynamic driving for non-equilibrium systems via mismatch elimination during DNA strand displacement,

    N. E. C. Haley, T. E. Ouldridge, I. Mullor Ruiz, A. Geraldini, A. A. Louis, J. Bath, and A. J. Turberfield, “Design of hidden thermodynamic driving for non-equilibrium systems via mismatch elimination during DNA strand displacement,”Nature Communications, vol. 11, p. 2562, May 2020

    work page 2020

  77. [77]

    In situ Generation of RNA Complexes for Synthetic Molecular Strand-Displacement Circuits in Autonomous Systems,

    W. Bae, G.-B. V. Stan, and T. E. Ouldridge, “In situ Generation of RNA Complexes for Synthetic Molecular Strand-Displacement Circuits in Autonomous Systems,”Nano Letters, vol. 21, pp. 265–271, Jan. 2021

    work page 2021

  78. [78]

    Cotranscriptionally encoded RNA strand displacement circuits,

    S. W. Schaffter and E. A. Strychalski, “Cotranscriptionally encoded RNA strand displacement circuits,”Science Advances, vol. 8, p. eabl4354, Mar. 2022

    work page 2022

  79. [79]

    Spurious transcription factor binding: Non-functional or genetically redundant?,

    M. Spivakov, “Spurious transcription factor binding: Non-functional or genetically redundant?,” BioEssays, vol. 36, pp. 798–806, Aug. 2014. 20

    work page 2014

  80. [80]

    The immune system, adaptation, and machine learning,

    J. Farmer, N. H. Packard, and A. S. Perelson, “The immune system, adaptation, and machine learning,”Physica D: Nonlinear Phenomena, vol. 22, pp. 187–204, Oct. 1986

    work page 1986

Showing first 80 references.