Pith. sign in

REVIEW 2 major objections 6 minor 36 references

Pattern formation within phenotype-structured chemotactic populations

T0 review · 2 major / 6 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read For populations whose members differ continuously in chemotactic sensitivity and attractant secretion, the switching rate between phenotypes decides which trait average sets the threshold for spatial pattern formation.

desk verdict Clean formal extension of Keller-Segel to phenotype-structured populations, but the rare-switching threshold rests on an initially uniform phenotype distribution and the paper only checks that special case. read the letter →

arxiv 2506.03389 v1 pith:O2Z7NAVE submitted 2025-06-03 q-bio.PE

Tommaso Lorenzi , Kevin J. Painter This is my paper
classification q-bio.PE MSC 35B3635K5792C1792B05
keywords phenotype-structuredPDEKeller-SegelpatternformationTuringinstabilityphenotypicheterogeneitychemotaxisphenotypeswitchingautoattraction
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

This paper asks whether phenotypic heterogeneity changes when a chemotactic population can self-organize into spatial clusters through autoattraction. It extends the classic Keller-Segel model to a phenotype-structured partial differential equation in which chemotactic sensitivity and attractant secretion vary continuously across a phenotype, with individuals switching between phenotypes at a rate β. Linear stability analysis shows that with negligible switching the critical condition involves the trait-averaged autoattraction potential ⟨αχ⟩, whereas with fast switching it involves the product of the trait averages ⟨α⟩⟨χ⟩. Because positive trait correlation makes ⟨αχ⟩ > ⟨α⟩⟨χ⟩ and negative correlation makes it smaller, the switching rate determines whether positively or negatively correlated populations aggregate more easily. This matters because real populations display exactly this kind of trait heterogeneity and correlation, from Dictyostelium signaling to bark-beetle pheromone production.

What carries the argument

The object doing the work is the phenotype-structured Keller-Segel system (2.1), a non-local advection-diffusion-reaction PDE with linear phenotype diffusion of rate β. Turing-type linear stability analysis around the uniform-in-space, uniform-in-phenotype steady state (B.3) or (C.2) reduces the dispersion relation to a quadratic (when growth is negligible) or a cubic (when growth is included). The instability conditions involve the three trait averages defined in (2.3)–(2.5): ⟨α⟩, ⟨χ⟩, and ⟨αχ⟩. The Chebyshev integral inequality is the identity that links the two asymptotic criteria, since for monotone traits positive correlation gives ⟨αχ⟩ > ⟨α⟩⟨χ⟩ and negative correlation gives the reverse; this ordering is what makes switching rate decisive.

What would settle it

Set β = 0 and initialise n0(x, y) concentrated in a narrow band of phenotypes (e.g., only high secretors and low sensors), perturb s, and check whether aggregation occurs exactly when condition (3.1) holds; since the uniform-in-phenotype steady state is not unique at β = 0, failure of (3.1) to predict the onset would indicate the analysis is an artifact of the singular limit.

Watch

Extended reading notes

Core claim

The central discovery is that in a phenotype-structured chemotaxis model, the rate of phenotype switching selects which statistical summary of the trait distributions controls the onset of self-organisation. In the limit of negligible switching, the linear-stability threshold is set by the trait-averaged autoattraction potential ⟨αχ⟩ = (1/|Y|)∫ α(y)χ(y) dy; in the fast-switching limit it is set by the product of the trait averages ⟨α⟩⟨χ⟩. Because these two quantities differ according to trait correlation (by the Chebyshev integral inequality), slow switching favours aggregation for positively correlated traits, while fast switching favours negatively correlated ones. Under population growth, negligible switching can additionally produce Turing-wave (oscillatory) instabilities for negatively correlated traits, and early patterns can decay because aggregates overshoot the carrying capacity and preferentially kill the most chemotactic phenotypes.

Load-bearing premise

The analysis assumes the population sits near a phenotype-uniform steady state and takes the negligible-switching limit by formally dropping phenotype diffusion, even though at zero switching that steady state is not unique, so the predicted threshold for slow switchers may miss the influence of the initial phenotype distribution.

Editorial extensions

If this is right

  • For a fixed mean sensitivity and secretion rate, positively correlated populations can aggregate at lower trait-averaged values when switching is rare, while negatively correlated populations need faster switching to reach the same ease of aggregation.
  • Variation in a single trait (only secretion or only sensitivity) leaves the threshold equal to the product of averages, so switching rate does not change the capacity for self-organisation in that case.
  • With logistic growth and negligible switching, the instability can be of Turing-wave type (oscillating in space and time) when traits are negatively correlated, while fast switching recovers stationary patterns akin to the unstructured model.
  • Patterns predicted by linear stability can decay on long timescales because aggregates exceed carrying capacity and lose the most chemotactic phenotypes; phenotype switching sustains clusters by replenishing those phenotypes.
  • Numerically, the critical switching rate β interpolates between the slow and fast thresholds, with the transition direction set by whether traits are positively or negatively correlated.

Reading between the lines

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

  • A testable consequence not drawn by the authors: if the initial phenotype distribution at β = 0 is not flat, the linear-stability criterion is likely to become distribution-dependent, not just average-dependent, since the uniform-in-phenotype steady state is not unique at β = 0.
  • The model suggests a cost-accounting rule for the population: negatively correlated traits should keep switching on to aggregate cheaply, while positively correlated traits can save switching energy; this could be probed experimentally by modulating phenotypic noise in bacteria.
  • Extending the analysis to growth-structured traits (go-or-grow) would probably soften the 'hidden cost' of aggregation, because slow-growing chemotactic phenotypes would not be penalised as strongly by density-dependent loss.
  • The dichotomy between ⟨αχ⟩ and ⟨α⟩⟨χ⟩ gives a practical diagnostic: comparing the critical α0χ0 under high versus low switching could be used to infer whether secretion and sensitivity are positively or negatively correlated in a real population.
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.

Referee Report

2 major / 6 minor

Summary. The paper studies a phenotype-structured extension of the Keller–Segel autoattraction model in which the density n(t,x,y) is structured by phenotype y ∈ Y with chemotactic sensitivity χ(y), attractant secretion rate α(y), and phenotype switching modelled as diffusion in y with rate β. Turing-type linear stability analysis is performed about the uniform-in-space, phenotype-homogeneous steady state in the limits of negligible (β→0) and fast (β→∞) switching, and for arbitrary β when χ(y) ≡ χ₀, both with negligible growth (Section 3) and with density-dependent growth (Section 4). The derived conditions (3.1)–(3.3), (4.1)–(4.2) show that the threshold is set by ⟨αχ⟩ when switching is rare and by ⟨α⟩⟨χ⟩ when switching is fast, and they reduce exactly to the unstructured Keller–Segel thresholds when traits do not vary. The paper concludes that positively correlated traits favour aggregation when switching is rare and negatively correlated traits favour aggregation when switching is frequent, and supports this with extensive numerical simulations, including growth-driven Turing-wave instabilities and long-time decay of patterns.

Significance. If the rare-switching threshold (3.1) were valid in full generality, the paper would deliver a clean, falsifiable prediction: the sign of trait correlation determines whether slow or fast switching lowers the aggregation threshold. The analysis is parameter-free, the reductions to the classical Keller–Segel conditions in Remarks 3.1 and 4.1 are exact, the fast-switching thresholds (3.2), (4.1) and the χ-constant thresholds (3.3), (4.2) follow from regular limits, and the numerical exploration is extensive, fully parameterised, and surfaces genuinely interesting phenomena (Turing-wave instabilities; transient patterns due to a growth-dependent cost of chemotaxis). The manuscript is thus a substantial contribution to phenotype-structured pattern formation, with the caveat, developed in the major comments, that the rare-switching results rest on a singular limit around a non-unique base state; the headline correlation-ordering prediction is therefore currently established only for phenotype-uniform initial conditions.

major comments (2)
  1. [§3.1, App. B.1, condition (3.1)] Condition (3.1) is derived by a formal β→0 limit around the phenotype-homogeneous steady state (B.3), but the limit is singular: at β=0 the uniform-in-space steady state of system (2.1) is any non-negative g(y) with ∫_Y g = ρ_m, as the paper itself notes for the growth case in footnote 2 of Section 4.2. For a generic frozen phenotype distribution g(y), the linearised chemotactic forcing is g(y)χ(y), so the instability threshold is ∫_Y α(y)χ(y)g(y) dy > ηD_n + (m²π²/L²)D_nD_s, which reduces to (3.1) only when g is uniform. Because the g-weighted average of αχ can lie on either side of both ⟨αχ⟩ and ⟨α⟩⟨χ⟩ — for example, with α(y)=2y, χ(y)=2(1−y) and g(y)=3ρ_m y² on Y=(0,1), the threshold is 0.6ρ_m while ⟨αχ⟩ρ_m = 2ρ_m/3 and ⟨α⟩⟨χ⟩ρ_m = ρ_m — the ordering conclusion of Section 3.2.3 (rare switching favours positively correlated traits) is not robust to initial phenotype heterogeneity. The order of limits matters: for every fixed β>0 the homogeneous steady state is unique and phenotype mixing eventually erases the initial profile, but the mixing timescale 1/β diverges as β→0, so the instability acts on the frozen initial profile. All simulations in Sections 3.2 and 4.2 start from n₀(x,y)≡1, so they verify only the uniform-g special case. Please scope condition (3.1) to perturbations of the phenotype-homogeneous state, add robustness tests with non-uniform n₀, or provide a β→0⁺ asymptotic analysis of the β>0 eigenvalue problem.
  2. [§4.1, App. C.1, eqs. (C.5)–(C.8)] The growth-case analysis inherits the singular-limit issue of condition (3.1). At β=0 the uniform-in-space steady state is any g(y) with ∫_Y g = κ (again footnote 2 of Section 4.2), yet the cubic (C.5) and the sufficient conditions (C.6)–(C.8) are derived by linearising around the phenotype-homogeneous state (C.2) and deleting β∂_yy in (C.3). The numerical eigenvalue computation in Fig. 4(b) and the resulting Turing-wave prediction for the (−,0) setting therefore apply only to perturbations of the phenotype-homogeneous β=0 state, while Section 4.2 states more generally that 'emerging patterns oscillate in both space and time' under (−,0). The restriction should be stated where these predictions are made, and the robustness of the Turing-wave instability to non-uniform frozen phenotype distributions should be tested numerically.
minor comments (6)
  1. [§4.2, footnote 2] The footnote correctly identifies the non-uniqueness of β=0 uniform steady states, but it appears only in the discussion of long-time behaviour; this non-uniqueness should also be acknowledged where conditions (3.1) and (C.5)–(C.8) are introduced, since it is the source of the major comments above.
  2. [Abstract] The phrase 'recent studies have shed illumination on the inherent heterogeneity' is unidiomatic; consider 'shed light on'.
  3. [§2.2, eq. (2.5)] The term 'autoattraction potential' is used in the Introduction for ⟨αχ⟩, but the quantity is first defined in (2.5) without the name; introduce the terminology at the definition.
  4. [§5, Refs. [16,31]] The Discussion cites 'binary phenotype' models via '[16, 31]', but [31] concerns a chemotaxis system with indirect signal production rather than a discrete phenotype model; please verify the citation.
  5. [Figs. 3 and 5] The captions of Figs. 3 and 5 should state that the numerical threshold interpolation is obtained with the phenotype-uniform perturbation (B.3)/(C.2) as initial data, making the special-case scope of the interpolation result explicit.
  6. [Appendices B–C] The numerical method behind the simulations in Figs. 2–5 (discretisation, time integration, tolerances) is not specified; adding these details would aid reproducibility.

Circularity Check

0 steps flagged · score 0.0 of 10

Self-contained derivation; no circularity found.

full rationale

The derivation chain is self-contained. The central patterning conditions (3.1), (3.2), (3.3), (4.1), and (4.2) are obtained by explicit linear stability analyses in Appendices B and C starting from the stated phenotype-structured PDE system (2.1), and they reduce to the unstructured Keller-Segel thresholds (A.4) and (A.5) when trait variation is absent (Remarks 3.1 and 4.1). No parameter is fitted to data and then renamed as a prediction; the thresholds are algebraic conditions derived from the model. The comparison between ⟨αχ⟩ and ⟨α⟩⟨χ⟩ is supported by the external Chebyshev integral inequality, not by a prior claim of the authors. Self-citations such as [16], [22], and [15] are contextual or supportive and are not load-bearing: the fast-switching reduction is re-derived within this paper in Appendix B.2 and Appendix C.2. The paper itself flags the genuine mathematical caveat that for β=0 the uniform-in-space steady state need not be phenotype-homogeneous (footnote 2, Section 4.2), which limits the generality of the β→0 predictions but is not a circular step. Overall, the derivation does not reduce to its inputs by construction, so the circularity score is 0.

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

No parameters are fitted to data; simulation values are illustrative. The model relies on standard PDE modeling assumptions and classical linear stability tools, plus the specific choice of linear phenotype diffusion. No new physical entities are introduced.

assumptions (5)
  • domain assumption Phenotype switching is modeled as linear diffusion beta d_yy n with constant beta.
    Section 2.1, equation (2.1)1; the entire analysis and the beta -> 0 and beta -> infinity limits depend on this additive diffusion structure.
  • domain assumption The unperturbed state for linear stability is homogeneous in both space and phenotype, n(y) = n*.
    Appendices B and C, ansatz (B.4); initial data n0(x,y) = 1 in the simulations. For beta = 0 this steady state is not unique.
  • standard math Chebyshev's integral inequality gives the ordering of <alpha chi> versus <alpha><chi> for monotone traits.
    Section 2.2, equations (2.10)-(2.12).
  • standard math Routh-Hurwitz conditions for a cubic with positive b coefficient determine when an eigenvalue has positive real part.
    Appendix C.1, analysis of equation (C.5).
  • standard math Turing-type linear stability analysis around a spatially uniform steady state predicts the onset of pattern formation.
    Appendix A and Appendix B; standard method, though the paper uses formal asymptotic limits.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Pattern formation within phenotype-structured chemotactic populations." pith.science (2026). https://pith.science/paper/O2Z7NAVE

@misc{pith2026250603389,
  author       = {Pith},
  title        = {Pith review of: Pattern formation within phenotype-structured chemotactic populations},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/O2Z7NAVE}},
  note         = {Machine review of arXiv:2506.03389}
}
read the original abstract

Populations can become spatially organised through chemotaxis autoattraction, wherein population members release their own chemoattractant. Standard models of this process usually assume phenotypic homogeneity, but recent studies have shed illumination on the inherent heterogeneity within populations: in terms of chemotactic behaviour, trait heterogeneity can range from the sensitivity to attractant gradients to the rate at which attractants are produced. We propose a framework that accounts for this heterogeneity, extending the standard Keller-Segel model to a non-local formulation in which the population is continuously structured across some phenotype state space. Focussing on autoattraction, we allow both the chemotactic sensitivity and the rate of attractant secretion to vary across the population and suppose members can switch between different phenotype states. We extend classical Turing-type linear stability analyses to determine the impact of phenotypic structuring on pattern formation, showing that the rate of switching influences both the critical condition for self-organisation and subsequent pattern dynamics. Scenarios in which the chemotactic sensitivity and attractant secretion are positively or negatively correlated are used to highlight the significance of these results.

Figures

Figures reproduced from arXiv: 2506.03389 by the authors.

Figure 1
Figure 1. Self-organisation via positive feedback of autoattraction and related case study scenarios. (a) Autoattraction in a phenotypically homogeneous population, where each member secretes its own chemoat￾tractant. (b) Autoattraction in a phenotypically heterogeneous population, where members vary with respect to their rate of attractant secretion and their chemotactic sensitivity. In this schematic the traits are negative… view at source ↗
Figure 2
Figure 2. Pattern formation in the structured chemotaxis model (2.1) under negligible population growth with no trait/single trait variation. (a) Predicted patterning region in (α0, χ0)-space for the structured chemotaxis model (2.1) with R ≡ 0, under no trait variation (i.e. χ(y) and α(y) defined via (2.6) and (2.8)), variation only in attractant secretion rate (i.e. χ(y) and α(y) defined via (2.6) and (2.9)), and variation … view at source ↗
Figure 3
Figure 3. Pattern formation in the structured chemotaxis model (2.1) under negligible population growth with twin trait variation. (a) Predicted patterning region in (α0, χ0)-space under fast (∞) or negligible (0) phenotype switching and positively (+) or negatively (−) correlated traits; α(y) and χ(y) are defined according to (2.13) and (2.14), respectively, with pα = pχ = 1. (b)-(c) Summary of results of numerical simulatio… view at source ↗
Figures from the paper (2 more)
Figure 4
Figure 4. Figure 4: Pattern formation in the structured chemotaxis model (2.1) under population growth with negligible or fast phenotype switching. (a) Predicted patterning region in (α0, χ0)-space for the structured chemotaxis model (2.1), with R(ρ) defined via (4.3), under the four corr…
Figure 5
Figure 5. Figure 5: Pattern formation in the structured chemotaxis model (2.1) under population growth with generic phenotype switching. (a) Summary of results of numerical simulations across (β, χ0)-space for positively (left) and negatively (right) correlated traits, showing short-term …

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

36 extracted references · 36 canonical work pages

  1. [16]

    F. R. Macfarlane, T. Lorenzi, and K. J. Painter , The impact of phenotypic heterogeneity on chemotactic self-organisation, Bulletin of Mathematical Biology, 84 (2022), p. 143. 25

    work page 2022

  2. [1]

    Ackermann, A functional perspective on phenotypic heterogeneity in microorganisms, Nature Reviews Microbiology, 13 (2015), pp

    M. Ackermann, A functional perspective on phenotypic heterogeneity in microorganisms, Nature Reviews Microbiology, 13 (2015), pp. 497–508

    work page 2015

  3. [2]

    Almeida, J

    L. Almeida, J. A. Denis, N. Ferrand, T. Lorenzi, A. Prunet, M. Sabbah, and C. Villa , Evolutionary dynamics of glucose-deprived cancer cells: insights from experimentally informed mathematical modelling, Journal of the Royal Society Interface, 21 (2024), p. 20230587

    work page 2024

  4. [3]

    Bellomo, A

    N. Bellomo, A. Bellouquid, Y. Tao, and M. Winkler, Toward a mathematical theory of Keller–Segel models of pattern formation in biological tissues , Mathematical Models and Methods in Applied Sciences, 25 (2015), pp. 1663–1763

    work page 2015

  5. [4]

    Carter and K

    B. Carter and K. Zhao , The epigenetic basis of cellular heterogeneity , Nature Reviews Genetics, 22 (2021), pp. 235–250

    work page 2021

  6. [5]

    R. H. Chisholm, T. Lorenzi, L. Desvillettes, and B. D. Hughes , Evolutionary dynamics of phenotype-structured populations: from individual-level mechanisms to population-level consequences , Zeitschrift f¨ ur angewandte Mathematik und Physik, 67 (2016), pp. 1–34

    work page 2016

  7. [6]

    G´enieys, V

    S. G´enieys, V. Volpert, and P. Auger , Adaptive dynamics: modelling Darwin ’s divergence principle, Comptes Rendus Biologies, 329 (2006), pp. 876–879

    work page 2006

  8. [7]

    J. M. Keegstra, F. Carrara, and R. Stocker , The ecological roles of bacterial chemotaxis , Nature Reviews Microbiology, 20 (2022), pp. 491–504

    work page 2022

Show all 36 references
  1. [8]

    E. F. Keller and L. A. Segel , Initiation of slime mold aggregation viewed as an instability , Journal of Theoretical Biology, 26 (1970), pp. 399–415

    1970

  2. [9]

    Keller and K

    L. Keller and K. Pantel, Unravelling tumour heterogeneity by single-cell profiling of circulating tumour cells, Nature Reviews Cancer, 19 (2019), pp. 553–567

    2019

  3. [10]

    A. L. Krause, E. A. Gaffney, T. J. Jewell, V. Klika, and B. J. W alker, Turing instabilities are not enough to ensure pattern formation , Bulletin of Mathematical Biology, 86 (2024), p. 21

    2024

  4. [11]

    A. L. Krause, E. A. Gaffney, P. K. Maini, and V. Klika , Modern perspectives on near-equilibrium analysis of turing systems , Philosophical Transactions of the Royal Society A, 379 (2021), p. 20200268

    2021

  5. [12]

    A. N. Landge, B. M. Jordan, X. Diego, and P. M ¨uller, Pattern formation mechanisms of self- organizing reaction-diffusion systems, Developmental Biology, 460 (2020), pp. 2–11

    2020

  6. [13]

    Lorenzi, F

    T. Lorenzi, F. R. Macfarlane, and K. J. Painter , Derivation and travelling wave analysis of phenotype-structured haptotaxis models of cancer invasion , European Journal of Applied Mathematics, 36 (2025), pp. 231–263

    2025

  7. [14]

    Lorenzi and K

    T. Lorenzi and K. J. Painter , Trade-offs between chemotaxis and proliferation shape the phenotypic structuring of invading waves , International Journal of Non-Linear Mechanics, 139 (2022), p. 103885

    2022

  8. [15]

    Lorenzi, K

    T. Lorenzi, K. J. Painter, and C. Villa, Phenotype structuring in collective cell migration: a tutorial of mathematical models and methods , Journal of Mathematical Biology, 90 (2025), p. 61

    2025

  9. [17]

    H. H. Mattingly and T. Emonet, Collective behavior and nongenetic inheritance allow bacterial popu- lations to adapt to changing environments , Proceedings of the National Academy of Sciences, 119 (2022), p. e2117377119

    2022

  10. [18]

    D. S. Mitrinovic and P. M. V asic, Analytic inequalities, vol. 1, Springer, 1970

    1970

  11. [19]

    B. Ni, R. Colin, H. Link, R. G. Endres, and V. Sourjik , Growth-rate dependent resource invest- ment in bacterial motile behavior quantitatively follows potential benefit of chemotaxis , Proceedings of the National Academy of Sciences, 117 (2020), pp. 595–601

    2020

  12. [20]

    K. J. Painter, Mathematical models for chemotaxis and their applications in self-organisation phenomena, Journal of Theoretical Biology, 481 (2019), pp. 162–182

    2019

  13. [21]

    K. J. Painter and T. Hillen , Spatio-temporal chaos in a chemotaxis model , Physica D: Nonlinear Phenomena, 240 (2011), pp. 363–375

    2011

  14. [22]

    K. J. Painter and M. Winkler , Phenotype switching in chemotaxis aggregation models controls the spontaneous emergence of large densities , SIAM Journal on Applied Mathematics, 83 (2023), pp. 2096– 2117

    2023

  15. [23]

    Perthame and S

    B. Perthame and S. G´enieys, Concentration in the nonlocal Fisher equation: the Hamilton-Jacobi limit , Mathematical Modeling of Natural Phenomena, 2 (2007), pp. 135–151

    2007

  16. [24]

    T. V. Phan, H. H. Mattingly, L. Vo, J. S. Marvin, L. L. Looger, and T. Emonet , Direct mea- surement of dynamic attractant gradients reveals breakdown of the Patlak–Keller–Segel chemotaxis model , Proceedings of the National Academy of Sciences, 121 (2024), p. e2309251121

    2024

  17. [25]

    D. S. Pureswaran, B. T. Sullivan, and M. P. Ayres, High individual variation in pheromone produc- tion by tree-killing bark beetles (Coleoptera: Curculionidae: Scolytinae) , Naturwissenschaften, 95 (2008), pp. 33–44

    2008

  18. [26]

    W. J. Ridgway, M. P. Dalwadi, P. Pearce, and S. J. Chapman , Motility-induced phase separation mediated by bacterial quorum sensing , Physical Review Letters, 131 (2023), p. 228302

    2023

  19. [27]

    Rietkerk, R

    M. Rietkerk, R. Bastiaansen, S. Banerjee, J. van de Koppel, M. Baudena, and A. Doelman , Evasion of tipping in complex systems through spatial pattern formation , Science, 374 (2021), p. eabj0359

    2021

  20. [28]

    M. M. Salek, F. Carrara, V. Fernandez, J. S. Guasto, and R. Stocker , Bacterial chemo- taxis in a microfluidic T-maze reveals strong phenotypic heterogeneity in chemotactic sensitivity , Nature Communications, 10 (2019), p. 1877

    2019

  21. [29]

    Schreiber and M

    F. Schreiber and M. Ackermann, Environmental drivers of metabolic heterogeneity in clonal microbial populations, Current Opinion in Biotechnology, 62 (2020), pp. 202–211

    2020

  22. [30]

    R. E. Stace, T. Stiehl, M. A. Chaplain, A. Marciniak-Czochra, and T. Lorenzi , Discrete and continuum phenotype-structured models for the evolution of cancer cell populations under chemotherapy , Mathematical Modelling of Natural Phenomena, 15 (2020), p. 14

    2020

  23. [31]

    Tao and H

    Y. Tao and H. Zhang , Nonlinear transmission exponent for boundedness of solutions to a chemotaxis system with indirect signal production , Applied Mathematics Letters, 149 (2024), p. 108928

    2024

  24. [32]

    Thiessen, M

    R. Thiessen, M. Conte, T. Stepien, and T. Hillen , Go-or-grow models in biology: a monster on a leash, arXiv preprint arXiv:2412.05191, (2024). 26

    2024

  25. [33]

    A. M. Turing , The chemical basis of morphogenesis , Philosophical Transactions of the Royal Society London: B, 252 (1952), pp. 37–72

    1952

  26. [34]

    Verhulst, Notice sur la loi que la population suit dans son accroissement , Correspondence Math- ematique et Physique, 10 (1838), pp

    P.-F. Verhulst, Notice sur la loi que la population suit dans son accroissement , Correspondence Math- ematique et Physique, 10 (1838), pp. 113–129

  27. [35]

    A. J. W aite, N. W. Frankel, and T. Emonet , Behavioral variability and phenotypic diversity in bacterial chemotaxis, Annual Review of Biophysics, 47 (2018), pp. 595–616

    2018

  28. [36]

    C. J. Weijer , Dictyostelium morphogenesis , Current Opinion in Genetics & Development, 14 (2004), pp. 392–398. 27

    2004

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.