Pith. sign in

REVIEW 2 major objections 5 minor 1 cited by

The Dimension of the Disguised Toric Locus of a Reaction Network

T0 review · 2 major / 5 minor · reviewed 2026-08-11 · deepseek-v4-flash

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

Pith's one-line read The paper states an exact dimension formula for the disguised toric locus of a reaction network, making the size of the 'hidden complex-balanced' parameter region computable from linear algebra.

desk verdict The dimension formula at the paper's heart is wrong as stated—signs of D0(G) and J0(G1) are reversed relative to the paper's own map—but the homeomorphism and the examples are worth a close look. read the letter →

arxiv 2412.02620 v1 pith:5WV46FIY submitted 2024-12-03 q-bio.MN math.DS

Gheorghe Craciun , Abhishek Deshpande , Jiaxin Jin This is my paper · ORCID
classification q-bio.MNmath.DS MSC 92C4237N2514M25
keywords disguisedtoriclocusmass-actionkineticscomplex-balancedsystemsreactionnetworksdynamicalequivalencedimensionformulastoichiometricsubspace
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

Complex-balanced reaction networks have very stable and predictable dynamics, but most reaction networks are not complex-balanced. This paper studies the disguised toric locus: the set of positive rate constants under which a network's mass-action system is dynamically equivalent to some complex-balanced system on a different network. The paper's central claim is an exact dimension formula: the dimension of this locus is the maximum, over weakly reversible subgraphs of the complete graph on the source vertices, of a sum of linear-algebraic dimensions. If correct, the formula turns a nonlinear dynamical question into linear algebra and shows that the disguised toric locus can be full-dimensional even when the ordinary toric locus is empty or measure-zero. The paper applies the formula to Thomas-type and circadian-clock models and finds that both have full-dimensional disguised toric loci.

What carries the argument

The load-bearing object is the map $\hat{\Psi}$ from the augmented set of realizable complex-balanced fluxes — the cone $J_R(G_1,G)$ together with the linear space $J_0(G_1)$ — and the invariant polyhedron of $G_1$, to the R-disguised toric locus $K_{R\text{-disg}}(G,G_1)$ together with the linear space $D_0(G)$. It sends a flux, a steady-state location, and the $D_0(G)$-projections of the rate vector to the unique rate vector producing those dynamics and its $J_0(G_1)$-projections. Proving that $\hat{\Psi}$ is a homeomorphism, via continuity of the inverse using an implicit-function theorem for steady states and a convergence argument for fluxes, is what converts the topological dimension of the locus into the sum of linear-algebraic dimensions.

What would settle it

Construct a minimal weakly reversible network with $\dim D_0(G)=1$ and $\dim J_0(G_1)=0$, parametrize $K_{R\text{-disg}}(G,G_1)$ directly from the defining equivalence equations, and compare its dimension with the formula; the paper's worked examples have both defect dimensions equal to zero, so they cannot distinguish the two sign conventions, while a nonzero case would settle the formula as printed.

Watch

Extended reading notes

Core claim

The discovery is Theorem 7.3: for a reaction network $G$, with complete graph $G_c$ on the source vertices, $$\dim(K_{R\text{-disg}}(G)) = \max_{G' \sqsubseteq G_c} \{ \dim(J_R(G',G)) + \dim(S_{G'}) + \dim(J_0(G')) - \dim(D_0(G)) \},$$ and $\dim(K_{\text{disg}}(G))$ is the same maximum restricted to subgraphs for which the positive locus is nonempty. The proof rests on a map $\hat{\Psi}$ defined in Definition 5.3 that pairs each rate vector in $K_{R\text{-disg}}(G,G_1)$ with its complex-balanced steady state and with projections onto the linear spaces $D_0(G)$ and $J_0(G_1)$; Theorem 6.10 shows this map is a homeomorphism. Because dimension is invariant under homeomorphism, the dimension of the pairwise locus follows by adding the dimensions of the domain factors. The examples then compute $\dim(K_{\text{disg}}(G))=6$ for the Thomas-type model and $7$ for the circadian-clock model, which are full-dimensional in the ambient parameter spaces even though their ordinary toric loci are empty or measure-zero.

Load-bearing premise

The formula assumes that the homeomorphism between the augmented parameter set and the product of the flux cone, invariant polyhedron, and linear defect spaces transfers dimensions factor by factor, with no hidden cancellation or overlap.

Editorial extensions

If this is right

  • The dimension of the disguised toric locus is computed from linear algebra alone: enumerate weakly reversible subgraphs of $G_c$, compute the cone dimension $\dim(J_R)$, the stoichiometric dimension, and the two defect dimensions, then take the maximum.
  • A network whose toric locus is empty or measure-zero can still have a full-dimensional disguised toric locus, so complex-balanced-style dynamics can persist on large open regions of parameter space.
  • For the Thomas-type and circadian-clock examples, the disguised toric locus has dimension equal to the number of reactions, meaning the parameter region is full-dimensional rather than a lower-dimensional subvariety.
  • When $K_{\text{disg}}(G,G_1)$ is nonempty, it has the same dimension as $K_{R\text{-disg}}(G,G_1)$, so allowing negative rate constants does not enlarge the dimension of the parameter region.

Reading between the lines

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

  • The formula suggests a direct computational pipeline: enumerate weakly reversible subgraphs, solve the linear feasibility problem defining $J_R$, and read off the dimension; the paper identifies efficient computation of $\dim(J_R)$ as the remaining bottleneck.
  • The two worked examples have $\dim D_0(G)=0$ and $\dim J_0(G_1)=0$, so neither tests the relative sign of the two defect terms in the printed formula; a small network with nonzero values for both would determine whether the theorem's sign convention or the example's sign convention is the one the homeomorphism actually yields.
  • The same homeomorphism strategy may carry over to other realization loci, such as detailed-balanced or weakly reversible deficiency-one realizations, because the proof only uses flux equivalence and the linear structure of $D_0$ and $J_0$.
  • One could test whether full-dimensionality is typical by sampling random networks and solving the linear feasibility problem; if it is, hidden complex-balanced structure may be common in biochemical models rather than exceptional.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

2 major / 5 minor

Summary. The paper studies the disguised toric locus Kdisg(G) and the R-disguised toric locus KR-disg(G) of a reaction network, i.e., the sets of rate constants for which the mass-action system is dynamically equivalent to a complex-balanced system on some weakly reversible subgraph of the complete graph Gc. It constructs a homeomorphism Ψ-hat (Definition 5.3) between a product built from complex-balanced flux systems and a product built from the R-disguised toric locus, and uses invariance of dimension to derive an exact dimension formula (Theorem 7.2), then extends it by taking maxima over weakly reversible subgraphs (Theorem 7.3). The paper applies these formulas to a Thomas-type model and a circadian clock model, claiming both disguised toric loci are full-dimensional.

Significance. The goal is valuable: replacing the previously known lower bounds on dim(KR-disg) and dim(Kdisg) with an exact formula would be a substantial contribution, and the construction of an explicit homeomorphism is an interesting technique that goes beyond the earlier lower-bound arguments. The paper includes a long, detailed proof of the continuity of the inverse map and a semialgebraic dimension framework. However, the central dimension formula contains a sign error that invalidates the theorem as stated; the examples in Section 8 use the corrected formula, which is why the error does not surface there. The homeomorphism construction and the surrounding lemmas appear to be a real asset, but the main result needs to be corrected before the paper's claims can be accepted.

major comments (2)
  1. [§7, Theorem 7.2(a), Eq. (7.3)] The sign in front of dim(D0(G)) and dim(J0(G1)) is reversed. By Definition 5.3 and the homeomorphism in Theorem 6.10, Ψ-hat maps JRhat(G1,G) × ((x0+SG1)∩R^n_>0) × R^b onto KR-disg(G,G1) × R^a, with b = dim D0(G) and a = dim J0(G1). Invariance of dimension therefore yields dim(KR-disg(G,G1)) + a = dim(JRhat(G1,G)) + dim(SG1) + b. Combining this with Lemma 7.1 gives dim(KR-disg(G,G1)) = dim(JR(G1,G)) + dim(SG1) + dim(D0(G)) - dim(J0(G1)). The proof's displayed equation, dim(KR-disg(G,G1)) + dim(D0(G)) = dim(JR(G1,G)) + dim(SG1) + dim(J0(G1)), places the R^b factor on the wrong side of the homeomorphism equation. Example 8.1 itself uses the corrected formula (+ dim(D0(G)) - dim(J0(G1))), so the example masks the error because both quantities vanish there.
  2. [§7, Theorem 7.3] Theorem 7.3 inherits the sign error from Theorem 7.2. The max formulas as printed state + dim(J0(G')) - dim(D0(G)), whereas the correct expression is + dim(D0(G)) - dim(J0(G')). Since dim(D0(G)) is constant in the max, the error changes the contribution of each candidate subgraph by 2 dim(J0(G')) and can change which subgraph attains the maximum; therefore the stated formulas for dim(KR-disg(G)) and dim(Kdisg(G)) are false whenever some nonzero J0(G') is involved. The central claim of the paper is thus incorrect as stated, even though the corrected formula appears in Example 8.1.
minor comments (5)
  1. [§6, proof of Theorem 6.10] The word 'homomorphism' in the final sentence should be 'homeomorphism'.
  2. [§5, Lemma 5.6] In the last paragraph of the proof, 'an orthonormal basis of the subspace J0(G)' should refer to J0(G1), since the vectors Ai form a basis of J0(G1).
  3. [§8, Examples 8.1 and 8.2] The assertions 'being R-realizable in G imposes no further constraints' and 'given any k1 ∈ K(G1) there exists k such that (G,k)∼(G1,k1)' are stated by graph inspection but are used to compute dim(JR(G1,G)); they should be justified explicitly or replaced by a verifiable construction.
  4. [§8, Example 8.1] The displayed inclusion writes a subset of R^7_>0 although the network G is stated to contain 6 reactions; the notation and the number of rate constants should be reconciled.
  5. [§7, Theorem 7.2] The use of invariance of dimension is applied to a convex cone and a semialgebraic set that are not manifolds; a brief justification (or a citation to semialgebraic dimension theory) that the homeomorphism preserves dimension on the dense open submanifold subsets would make the proof more rigorous.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the dimension formula is derived from a homeomorphism proved in this paper, and the same-group citations are concrete checkable lemmas; the reversed D0/J0 signs in Eq. (7.3) are a correctness defect, not a circular step.

full rationale

The paper's derivation chain is not circular. Theorem 7.3's dimension formula is obtained by applying invariance of dimension to the map Psi-hat (Definition 5.3), which this paper constructs and proves to be a homeomorphism in Sections 5-6 (Lemmas 5.4-5.7 and 6.2-6.9, Theorems 6.9-6.10); that proof is new here and does not assume the formula. Each input to the formula - dim(JR(G1,G)), dim(SG1), dim(J0(G1)), dim(D0(G)) - is defined independently of the output dim(KR-disg(G,G1)) (Definitions 2.10, 3.3, 4.7, Lemma 4.8) and is computed directly from the graphs, so the result is not a redefinition, and no parameter is fitted and then renamed a prediction; the examples further corroborate their full-dimensional conclusions with independent matrix-tree-theorem computations of explicit subsets of Kdisg(G). The paper does lean on same-group prior work ([23] Lemmas 2.11, 3.4, 4.8; [34] Proposition 3.5 and the G' being a weakly reversible subgraph of Gc reduction; [5] toric-locus facts), but those are parameter-free lemmas whose stated assumptions do not include the target dimension formula, and the uniqueness inputs (Theorems 2.6-2.7) are external (Horn-Jackson; Johnston), so the self-citations are real evidence rather than a load-bearing self-citation chain. Section 9 is candid that dim(JR(G',G)) remains to be characterized (future work [45]), confirming the paper reduces the hard computation instead of presupposing it. One in-scope issue is a correctness defect, not circularity: the proof of Theorem 7.2 displays dim(KR-disg) + dim(D0(G)) = dim(JR) + dim(SG1) + dim(J0(G1)), whereas the homeomorphism of Definition 5.3 (domain ... x R^b, codomain ... x R^a, with b = dim D0(G) and a = dim J0(G1)) forces dim(KR-disg) + dim(J0(G1)) = dim(JR) + dim(SG1) + dim(D0(G)); Eq. (7.3) accordingly has the D0/J0 signs reversed relative to the construction and relative to the formula displayed in Example 8.1, which matches only because both quantities are zero there. This internal inconsistency is for a correctness pass, not a circularity claim.

Assumptions & free parameters 0 free parameters · 5 assumptions · 0 invented entities

No fitted numerical parameters or new physical entities. The main unrecognized input is the dimension-counting principle in Theorem 7.2, which is applied with the wrong sign. The examples rely on unproved graph-inspection claims.

assumptions (5)
  • standard math Invariance of dimension applies to the homeomorphic semialgebraic sets in Theorem 7.2.
    Used in the proof of Theorem 7.2 to transfer dimension from the domain of Psi-hat to KR-disg x R^a. The factors R^a and R^b are assigned to the wrong sides, producing the sign error.
  • standard math The implicit function theorem yields a local continuous map from toric rate constants to their unique steady states (Lemma 6.6).
    Needed for the continuity of the inverse map Psi-hat^{-1} in Section 6.
  • domain assumption Theorems 2.6 and 2.7 on uniqueness of positive steady states and the Jacobian kernel for complex-balanced systems.
    Imported from the reaction network literature and used throughout the proof.
  • standard math The dimension of a semialgebraic set is the maximum dimension of its locally submanifold points, and the dimension of a finite union is the maximum (Lemma 4.4, Remark 4.5).
    Justifies the max over subgraphs in Theorem 7.3.
  • ad hoc to paper In the examples, 'being R-realizable in G imposes no further constraints' and 'given any k1 there exists k' are asserted by graph inspection.
    These assumptions are needed for the claimed dimensions of the Thomas and circadian examples; the paper does not prove them in detail.

how reviews work

0 comments
Cite this review

Pith. "Pith review of The Dimension of the Disguised Toric Locus of a Reaction Network." pith.science (2026). https://pith.science/paper/5WV46FIY

@misc{pith2026241202620,
  author       = {Pith},
  title        = {Pith review of: The Dimension of the Disguised Toric Locus of a Reaction Network},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/5WV46FIY}},
  note         = {Machine review of arXiv:2412.02620}
}
abstract

Under mass-action kinetics, complex-balanced systems emerge from biochemical reaction networks and exhibit stable and predictable dynamics. For a reaction network $G$, the associated dynamical system is called $\textit{disguised toric}$ if it can yield a complex-balanced realization on a possibly different network $G_1$. This concept extends the robust properties of toric systems to those that are not inherently toric. In this work, we study the $\textit{disguised toric locus}$ of a reaction network - i.e., the set of positive rate constants that make the corresponding mass-action system disguised toric. Our primary focus is to compute the exact dimension of this locus. We subsequently apply our results to Thomas-type and circadian clock models.

Figures

Figures reproduced from arXiv: 2412.02620 by the authors.

Figure 1
Figure 1. (a) The E-graph G represents a Thomas-type model, with all edges labeled by the reaction rate constants k. (b) The E-graph G1 is weakly reversible, with all edges labeled by the reaction rate constants k1. The mass-action system (G1, k1) is complex-balanced. Since G is not weakly reversible, the system (G, k) is not complex-balanced, so classi￾cal complex-balanced theory offers limited insight into the dynamics of (… view at source ↗
Figure 2
Figure 2. (a) An E-graph with two reactions. The stoichiometric subspace corresponding to this graph is R 2 . (b) A weakly reversible E-graph. (c) A directed complete E-graph with three vertices. Note that the E-graph in (b) is a weakly reversible subgraph of the E-graph in (c). Definition 2.3 ([26, 27, 28, 29, 4, 30]). Consider an E-graph G = (V, E). Let ky→y′ denote the reaction rate constant corresponding to the reaction y… view at source ↗
Figure 3
Figure 3. (a) The E-graph G represents a Thomas-type model, where U denotes uric acid and V denotes oxygen. (b) The E-graph G1 is a weakly reversible subgraph of the complete graph formed by the source vertices of G. The graph G contains 6 reactions, hence Kdisg(G) ⊆ R 6 >0 . We claim that dim(Kdisg(G)) = 6. This implies that the disguised toric locus corresponding to G is of positive measure. To prove this claim, we consider… view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: (a) The E-graph G represents a circadian clock model. (b) The E-graph G1 is a weakly reversible subgraph of the complete graph formed by the source vertices of G. Note that the graph G contains 7 reactions, so Kdisg(G) ⊆ R 7 >0 . We claim that dim(Kdisg(G)) = 7. This i…

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Weakly reversible deficiency zero realizations of reaction networks

    q-bio.MN 2025-02 conditional novelty 6.0 of 10

    If a reaction network has a weakly reversible deficiency zero realization for all rate constants, that realization is unique and can be found by an algorithm.

Reference graph

Works this paper leans on

46 extracted references · 44 canonical work pages · cited by 1 Pith paper

  1. [23]

    A Lower Bound on the Dimension of the $\mathbb{R}$-Disguised Toric Locus of a Reaction Network

    G. Craciun, A. Deshpande, and J. Jin. A lower bound on the dimension of the R- disguised toric locus of a reaction network. https://arxiv.org/pdf/2305.00299, 2023

    work page Pith review arXiv 2023

  2. [1]

    Craciun and A

    G. Craciun and A. Deshpande. Homeostasis and injectivit y: a reaction network per- spective. J. Math. Biol. , 85(6):67, 2022

    work page 2022

  3. [2]

    Deshpande and M

    A. Deshpande and M. Gopalkrishnan. Autocatalysis in rea ction networks. Bull. Math. Biol., 76(10):2570–2595, 2014

    work page 2014

  4. [3]

    Ilyashenko

    Y. Ilyashenko. Centennial history of Hilbert’s 16th pro blem. Bulletin of the American Mathematical Society, 39(03):301–355, 2002

    work page 2002

  5. [4]

    Yu and G

    P. Yu and G. Craciun. Mathematical Analysis of Chemical R eaction Systems. Isr. J. Chem., 58(6-7):733–741, 2018

    work page 2018

  6. [5]

    Craciun, A

    G. Craciun, A. Dickenstein, A. Shiu, and B. Sturmfels. To ric dynamical systems. J. Symbolic Comput. , 44(11):1551–1565, 2009

    work page 2009

  7. [6]

    Dickenstein

    A. Dickenstein. Algebraic geometry tools in systems bio logy. Not. Am. Math. Soc , 67:1706–1715, 2020

    work page 2020

  8. [7]

    Horn and R

    F. Horn and R. Jackson. General mass action kinetics. Arch. Ration. Mech. Anal. , 47(2):81–116, 1972

    work page 1972

Show all 46 references
  1. [8]

    Anderson

    D. Anderson. A proof of the global attractor conjecture i n the single linkage class case. SIAM J. Appl. Math. , 71(4):1487–1508, 2011

    2011

  2. [9]

    Gopalkrishnan, E

    M. Gopalkrishnan, E. Miller, and A. Shiu. A geometric app roach to the global attrac- tor conjecture. SIAM J. Appl. Dyn. Syst. , 13(2):758–797, 2014. 40

    2014

  3. [10]

    C. Pantea. On the persistence and global stability of ma ss-action systems. SIAM J. Math. Anal. , 44(3):1636–1673, 2012

    2012

  4. [11]

    Craciun, F

    G. Craciun, F. Nazarov, and C. Pantea. Persistence and p ermanence of mass-action and power-law dynamical systems. SIAM J. Appl. Math. , 73(1):305–329, 2013

    2013

  5. [12]

    Boros and J

    B. Boros and J. Hofbauer. Permanence of weakly reversib le mass-action systems with a single linkage class. SIAM J. Appl. Dyn. Syst. , 19(1):352–365, 2020

    2020

  6. [13]

    G. Craciun. Toric differential inclusions and a proof of t he global attractor conjecture. arXiv preprint arXiv:1501.02860 , 2015

    2015 arXiv

  7. [14]

    Craciun and C

    G. Craciun and C. Pantea. Identifiability of chemical re action networks. J. Math. Chem., 44(1):244–259, 2008

    2008

  8. [15]

    Craciun, A

    G. Craciun, A. Deshpande, and J. Jin. Weakly reversible deficiency one realizations of reaction networks: an algorithmic perspective. Discrete and Continuous Dynamical Systems - B , 29(6):2786–2816, 2024

    2024

  9. [16]

    Craciun, A

    G. Craciun, A. Deshpande, and J. Jin. Weakly reversible realizations that obey the deficiency one theorem: an algorithmic perspective. In preparation, 2024

    2024

  10. [17]

    Boros, G

    B. Boros, G. Craciun, and P. Yu. Weakly reversible mass- action systems with infinitely many positive steady states. SIAM J. Appl. Math. , 80(4):1936–1946, 2020

    1936

  11. [18]

    Kothari and A

    S. Kothari and A. Deshpande. Endotactic and strongly en dotactic networks with infinitely many positive steady states. J. Math. Chem. , pages 1–25, 2024

    2024

  12. [19]

    Kothari, J

    S. Kothari, J. Jin, and A. Deshpande. Realizations thro ugh weakly reversible networks and the globally attracting locus. arXiv preprint arXiv:2409.04802 , 2024

    2024 arXiv

  13. [20]

    Moncus ´ ı, G

    L. Moncus ´ ı, G. Craciun, and M. Sorea. Disguised toric d ynamical systems. J. Pure and Appl. Alg. , 226(8):107035, 2022

    2022

  14. [21]

    Haque, M

    S. Haque, M. Satriano, M. Sorea, and P. Yu. The disguised toric locus and affine equivalence of reaction networks. SIAM J. Appl. Dyn. Sys. , 22(2):1423–1444, 2023

    2023

  15. [22]

    Craciun, A

    G. Craciun, A. Deshpande, and J. Jin. On the connectivit y of the disguised toric locus of a reaction network. J. Math. Chem. , 62(2):386–405, 2024

    2024

  16. [24]

    G. Craciun. Polynomial dynamical systems, reaction ne tworks, and toric differential inclusions. SIAM J. Appl. Algebra Geom. , 3(1):87–106, 2019. 41

    2019

  17. [25]

    Craciun and A

    G. Craciun and A. Deshpande. Endotactic networks and to ric differential inclusions. SIAM J. Appl. Dyn. Syst. , 19(3):1798–1822, 2020

    2020

  18. [26]

    Adleman, M

    L. Adleman, M. Gopalkrishnan, M. Huang, P. Moisset, and D. Reishus. On the mathematics of the law of mass action. In A Systems Theoretic Approach to Systems and Synthetic Biology I: Models and System Characterizations , pages 3–46. Springer, 2014

    2014

  19. [27]

    Waage and M

    P. Waage and M. Gulberg. Studies concerning affinity. J. Chem. Educ. , 63(12):1044, 1986

    1986

  20. [28]

    E. Voit, H. Martens, and S. Omholt. 150 years of the mass a ction law. PLOS Comput. Biol., 11(1):e1004012, 2015

    2015

  21. [29]

    Gunawardena

    J. Gunawardena. Chemical reaction network theory for i n-silico biologists. Notes available for download at http://vcp. med. harvard. edu/pa pers/crnt. pdf, 2003

    2003

  22. [30]

    Feinberg

    M. Feinberg. Lectures on chemical reaction networks. Notes of lectures given at the Mathematics Research Center, University of Wisconsin , page 49, 1979

    1979

  23. [31]

    E. Sontag. Structure and stability of certain chemical networks and applications to the kinetic proofreading model of t-cell receptor signal tr ansduction. IEEE Trans. Automat., 46(7):1028–1047, 2001

    2001

  24. [32]

    Topics in chemical reaction network theory

    Matthew Johnston. Topics in chemical reaction network theory. UWSpace, 2012

    2012

  25. [33]

    Deshpande

    A. Deshpande. Source-only realizations, weakly rever sible deficiency one networks, and dynamical equivalence. SIAM J. Appl. Dyn. , 22(2):1502–1521, 2023

    2023

  26. [34]

    Craciun, J

    G. Craciun, J. Jin, and P. Yu. An efficient characterizati on of complex-balanced, detailed-balanced, and weakly reversible systems. SIAM J. Appl. Math. , 80(1):183– 205, 2020

    2020

  27. [35]

    M. Coste. An introduction to semialgebraic geometry, 2 000

  28. [36]

    J. Lee. Introduction to topological manifolds, volume 202. Springer Science & Business Media, 2010

    2010

  29. [37]

    Guillemin and A

    V. Guillemin and A. Pollack. Differential topology, volume 370. American Mathemat- ical Soc., 2010

    2010

  30. [38]

    A. Hatcher. Algebraic topology. Cambridge University Press, 2005

    2005

  31. [39]

    J. Munkres. Elements of algebraic topology . CRC press, 2018

    2018

  32. [40]

    Basu and B

    S. Basu and B. Mishra. Computational and Quantitative Real Algebraic Geometry , volume 38. CRC Press, 2017. 42

    2017

  33. [41]

    Lairez and M

    P. Lairez and M. Safey El Din. Computing the dimension of real algebraic sets. In Proceedings of the 2021 on International Symposium on Symbol ic and Algebraic Computation, pages 257–264, 2021

    2021

  34. [42]

    J. Murray. Mathematical biology: I. An introduction , volume 17. Springer Science & Business Media, 2007

    2007

  35. [43]

    Leloup and A

    J. Leloup and A. Goldbeter. Chaos and birhythmicity in a model for circadian os- cillations of the per and tim proteins in drosophila. J. Theor. Biol. , 198(3):445–459, 1999

    1999

  36. [44]

    Craciun, J

    G. Craciun, J. Jin, and M. Sorea. The structure of the mod uli spaces of toric dynamical systems. arXiv preprint arXiv:2008.11468 , 2023

    2008 arXiv

  37. [45]

    Craciun, A

    G. Craciun, A. Deshpande, and J. Jin. On the relationshi p between dynamical equiv- alence and complex balancing. In preparation, 2023

    2023

  38. [46]

    Craciun, J

    G. Craciun, J. Jin, and M. Sorea. The toric locus of a reac tion network is a smooth manifold. https://arxiv.org/abs/2309.15241, 2023. 43

    2023 arXiv

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.