{"id":"0bbe0333-b1e5-455a-9932-8e7d434cfa74","arxiv_id":"2502.09554","paper_version":1,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":6.0,"correctness_risk":"high","formal_verification":"none","parameter_count":5,"one_line_summary":"Activity in confluent epithelia renormalizes the core energy of hexatic dislocations, producing a Kosterlitz-Thouless-Halperin-Nelson-Young type melting transition whose correlation-length exponent can range up to 1 depending on T1 and T2 topological rearrangements.","lead":"This paper proposes that the onset of collective migration in epithelial layers is a defect-driven melting transition, where cell activity lowers the energy cost of topological rearrangements and changes the critical behavior. It combines cell-resolved simulations with a renormalization-group analysis to predict a tunable correlation-length exponent that depends on whether the layer is closed or open to cell entry and exit.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Open-system exponent nu_bar=0.18 is ~5 sigma below the lower bound ~0.48 of Eq. (5) computed with the paper's own constants; with c1,c2,c3 treated as free parameters, the claimed exponent spectrum is not actually tested.","rationale":"In good faith, the qualitative picture of activity-driven KTHNY melting is plausible, and the simulation data do show the expected phase sequence. However, the paper's quantitative claim about a spectrum of exponents is the load-bearing part of the central argument. The reader's weakest assumption points to the equilibrium-like postulate, which is indeed fundamental but harder to settle with one decisive check. I find a more direct and internal problem: the open-system exponent is inconsistent with Eq. (5) when evaluated with the paper's own quoted constants, and the authors do not provide the measured c1, c2, c3 for their cell model that would resolve the mismatch. This makes the predicted range effectively unfalsifiable unless the constants are pinned down. The concrete test above would settle the issue: if the measured constants place the lower bound below 0.18, the fit is inside the predicted spectrum and the central claim survives; if not, Eq. (5) is not supported. Because the concern is specific and addressable, the existing CONDITIONAL verdict remains appropriate; no adjustment is needed.","tokens_in":19567,"tokens_out":6642,"duration_ms":92945,"concrete_test":"Measure the bare constants c1, c2, c3 for the MPF model in the open-system simulations: from the sub-threshold solid, count the statistical weights of configurations with one, two, and three dislocations (or of T1 and T2 events) and fit their l-dependence to Eqs. (S21) to extract c1, c2, c3. Then recompute the lower bound 0.5*(1 - c3/sqrt(c3^2 + 4*c1*c2)) and compare it with nu_bar = 0.18 +/- 0.06 from Fig. 2e. If the measured bound is above 0.18, Eq. (5) is contradicted; if below, report the measured c-values so the comparison is reproducible.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central quantitative claim is the 'well-defined spectrum' of critical exponents in Eq. (5). The open-system fit gives nu_bar=0.18 +/- 0.06 (Fig. 2e). Using the constants quoted in the paper for the triangular-lattice point-particle gas, c1=32*pi, c2=2597.84, c3=38.93, the lower bound of Eq. (5) is 0.5*(1 - c3/sqrt(c3^2 + 4*c1*c2)) = 0.48. The measured value is therefore about five standard deviations below the predicted range, yet the text describes it as 'close to the lower bound.' The stated escape is that c1, c2, and c3 'are expected to shift away from the classic estimate' and are treated as material parameters. That makes the spectrum non-predictive as stated: any fitted nu_bar in (0,1] can be rationalized post hoc by choosing different c-values. Unless the actual c1, c2, c3 for the multiphase-field model are measured and the lower bound recomputed, the central claim of a well-defined, material-dependent exponent range is not supported by the simulations. This is an internal quantitative mismatch, not merely a disagreement with equilibrium KTHNY expectations.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript proposes that the onset of collective migration in confluent epithelia is an activity-driven KTHNY-type melting transition. It argues that active stresses renormalize the dislocation core energy (Eq. 2), that confluency restricts the available dislocation-unbinding modes (leading to c3=0 for closed layers), and that open layers with both T1 and T2 processes yield a material-dependent exponent ν̄ in the range of Eq. (5). The RG analysis is paired with multiphase-field simulations of 1085 cells in closed and open systems, which show a solid-hexatic-liquid phase sequence and an exponentially divergent correlation length across the solid/hexatic boundary. The fitted exponents are ν̄=1.0±0.1 (closed) and ν̄=0.18±0.06 (open). A limiting case with only T2 processes (c2=0) is predicted to show power-law divergence of the correlation length.","tokens_in":19843,"tokens_out":7809,"duration_ms":73591,"significance":"If the quantitative claims held, the paper would establish a concrete and testable connection between topological rearrangements in epithelia and defect-mediated melting, and it would identify T1/T2 availability as a tuning parameter for critical behavior. The numerical evidence for the two-step melting sequence is credible: the authors use three independent realizations, long averaging, and multiple diagnostics (structure factor, orientational correlations, MSD, defect densities), and the analytical flow solution in Supplementary Sec. S1 is a useful standalone contribution. However, the central quantitative claim—the 'well-defined spectrum' of exponents in Eq. (5)—is not actually supported by the simulations as presented. The open-system fitted exponent falls well outside the spectrum computed with the paper's own constants, and the stated escape that the constants are material-dependent makes the spectrum non-predictive without an independent measurement. The equilibrium assumption at criticality is also not directly tested. If these issues are addressed, the work could make a solid contribution; in its present form the central quantitative claim needs substantial revision.","major_comments":[{"comment":"The open-system fit gives ν̄=0.18±0.06, while the lower bound of Eq. (5), computed with the constants stated in the text and in Supplementary Sec. S2 (c1=32π, c2≈2597.84, c3≈38.93), is 0.5[1−c3/(c3²+4c1c2)^1/2]≈0.48. The fitted value is therefore about five standard deviations below the claimed spectrum, not 'close to the lower bound' as stated in the text. Because the manuscript permits c1,c2,c3 to 'shift away from the classic estimate' and treats them as material parameters, the comparison is not a test of Eq. (5); any ν̄ in (0,1] could be rationalized by an appropriate choice of the c's. I request either an independent measurement of c1,c2,c3 for the multiphase-field model, or an explicit statement that the open-system exponent is outside the predicted range, with a discussion of what this implies for the theory.","section":"Eq. (5) and Fig. 2e"},{"comment":"The RG flow (Eqs. 3 and S21) and the dislocation Hamiltonian (Eq. S19) are imported from equilibrium KTHNY theory with a modified core energy; they are not derived from the active multiphase-field dynamics. The paper assumes that the transition is effectively equilibrium-like at criticality, but the a posteriori evidence—the phase sequence and an exponentially divergent correlation length—does not test whether the fluctuations obey Gibbs-Boltzmann statistics. I ask for a direct diagnostic, such as comparing the measured dislocation pair distribution or static structure factor with the equilibrium predictions of Eq. (S19), or measuring an effective temperature via a fluctuation-dissipation ratio. Without such a check, the central mechanism remains an assumption.","section":"Introduction and Supplementary Sec. S2"},{"comment":"The mapping between T2 processes and the constant c3 is only qualitative. In the simulations, T2 processes are implemented as random cell extrusion and intrusion events, but neither the rate nor the associated fugacity y0 (or effective core energy) is measured, and c3 is never extracted. Consequently, the claim that open systems 'render melting possible even at low activities' and thereby place ν̄ near the lower bound is not quantitatively connected to Eq. (5). The paper should either report measured T1/T2 rates and relate them to y0, or explicitly treat c3 as a free parameter and show that the fit remains inside the corresponding range.","section":"Main text after Eq. (5); Supplementary Sec. S3B"},{"comment":"The estimate of ξ+ from the defect density via n5-7∼ξ+^{-2}, and the subsequent fit to Eq. (4), are not described in sufficient detail. The number of data points, the fitting range in v0, the weighting of the error bars, and the goodness-of-fit are absent, and the text does not state whether alternative functional forms (e.g., a power law or a pure KT form) were tested. Given that the quantitative claim rests entirely on this fit, please provide these details and, if possible, an independent measure of ξ+ from the decay of g6(r) or S(q).","section":"Fig. 2e and Methods"}],"minor_comments":[{"comment":"Typography and language: 'loosing confluency' should be 'losing confluency'; 'KTNHY' in the Fig. 1a caption should be 'KTHNY'; 'multiphase filed' should be 'multiphase field'; and 'experiementally' in the main text should be 'experimentally'.","section":"Abstract and Fig. 1a caption"},{"comment":"The same symbol σ is used for the stress tensor and for its magnitude, which makes the derivation hard to follow. Please use a distinct symbol for the tensor (e.g., Σ) and define the scalar explicitly.","section":"Eq. (1)"},{"comment":"The sentence describing the quadrupole should read 'two +1/6 and two −1/6 disclinations' rather than 'two +1/6 and two +1/6 disclinations'.","section":"Supplementary Sec. S1"},{"comment":"The relation between the coarse-grained active stress σ(r) in Eq. (S49) and the scalar σ in Eq. (1) of the main text is not stated. Please explain explicitly how σ∼v0 follows from the simulation parameters.","section":"Supplementary Sec. S3A and Eq. (S49)"},{"comment":"The statement that Gaussian white noise introduced via a Langevin or Monte Carlo thermostat is 'indistinguishable from the thermal noise characteristic of Gibbs ensembles' is too strong; a Langevin thermostat does not automatically guarantee Gibbs statistics for the configurational degrees of freedom in an active system.","section":"Main text, paragraph on thermal-like fluctuations"},{"comment":"Axis labels and units for the horizontal axis are missing. The caption should specify the self-propulsion speed in units of a Dr, where a is the lattice spacing and Dr the rotational diffusivity.","section":"Fig. 2e"}],"recommendation":"major_revision","confidential_remarks":"The paper has a strong qualitative core and is likely to interest soft-matter and biological-physics readers. My main concern is the internal mismatch between Eq. (5) and the open-system exponent fit; this is a load-bearing quantitative claim that must be resolved before publication. The equilibrium assumption also needs a more direct test than the phase-sequence evidence. I do not think rejection is warranted at this stage, because the phase-sequence results, the active-flow solution, and the RG framework are valuable and can be revised. However, if the authors cannot provide an independent determination of c2 and c3 for their model, or cannot otherwise reconcile the fitted exponent with Eq. (5), the quantitative central claim should be substantially weakened."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Quick take on arXiv:2502.09554. The paper is worth a serious look: it adapts KTHNY melting theory to confluent epithelia, with activity renormalizing the dislocation core energy and confluency cutting the number of unbinding modes, and it backs the picture with multiphase-field simulations showing the expected solid-hexatic-liquid sequence and an exponentially divergent correlation length. The closed vs open tissue comparison, mapping T1/T2 processes onto c2/c3, is a genuinely nice idea.\n\nThe soft spot is the quantitative claim. The paper’s Eq. (5) gives a lower bound of 0.48 for the correlation-length exponent using the classic constants it quotes (c1=32π, c2≈2597.84, c3≈38.93). The open-system fit returns 0.18±0.06. That is about five standard deviations below the bound, not 'close to the lower bound' as the text says. The stated escape—that c1, c2, c3 are material parameters that will shift—means the spectrum is not actually predictive as written. Any value in (0,1] can be rationalized post hoc by choosing different c’s. To make the claim stand, the authors need to measure or estimate the actual c-values for the multiphase-field model, or present the exponent spectrum as a qualitative prediction with the constants left unspecified.\n\nThere is also a load-bearing assumption, acknowledged in the introduction: at criticality the transition is effectively equilibrium-like. The RG machinery and the interpretation of the fugacity all presuppose Gibbs-Boltzmann statistics. The only a posteriori check is the phase sequence, which does not test fluctuation spectra or response functions. That is a real gap, though not a fatal one if it is framed as a modelling assumption.\n\nWhat is genuinely good: the simulations are careful, the three-phase sequence is convincing, and the closed-system exponent 1.0 sits at the predicted upper bound. The paper is honest about its assumptions and cites the relevant literature.\n\nThis is for people working on active matter and epithelial mechanics. It deserves a proper peer review, but the referee should push hard on the exponent mismatch and on what would falsify the spectrum. My recommendation: send it to review with a request for major revision that either reconciles the measured exponent with the theoretical bound or explicitly weakens the claim.","headline":"A promising KTHNY-based theory of activity-driven epithelial melting, but the central exponent-spectrum claim is undercut by an internal numerical mismatch and needs revision before it can be taken as quantitative.","tokens_in":20416,"tokens_out":2673,"would_cite":false,"duration_ms":25344,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["82B26","82B27","82B28"],"pacs":[],"model":"deepseek-v4-flash","headline":"Active stress turns the onset of collective migration into a defect-driven melting transition.","keywords":["KTHNY melting","confluent epithelia","collective cell migration","T1 and T2 rearrangements","hexatic order","dislocation core energy","active matter","renormalization group"],"falsifier":"Measure the density of unbound 5–7 dislocations and the positional correlation length while independently varying the self-propulsion speed and the rate of cell intrusion/extrusion, and check whether the fitted exponent falls in the predicted range and whether the divergence is exponential across the solid–hexatic boundary. More directly, test the equilibrium assumption by measuring the fluctuation–dissipation ratio near criticality: if the effective temperature extracted from that ratio does not match the value used in the Boltzmann fugacity, the equilibrium mapping fails.","tokens_in":2104,"feed_emoji":"🧫","tokens_out":3881,"duration_ms":83713,"temperature":0.7,"pith_summary":"The paper argues that the onset of collective migration in a confluent epithelium is an activity-driven version of the two-dimensional Kosterlitz–Thouless–Halperin–Nelson–Young (KTHNY) melting scenario. In this picture active stresses produced by T1 cell intercalation reduce the effective energetic cost of creating 5–7 dislocations, so the dislocation fugacity becomes $y=\\exp[-(\\epsilon_c+\\sigma)a^2/(k_B T)]$. The result is an exponentially divergent correlation length across the solid–hexatic boundary, with an exponent $\\bar\\nu$ that is not universal but lies in a material-dependent range bounded by $1/2$ and $1$ (and whose lower limit depends on the ratio of T1 to T2 processes). This matters because it converts a qualitative question—when does an epithelium start to migrate?—into a quantitative scaling law that could be compared directly with experiments, and it predicts that tissues with different rates of cell division and extrusion should exhibit measurably different critical behavior.","feed_headline":"Epithelial migration is a tunable defect-driven melting","feed_subtitle":"Closed and open tissues land at different ends of a predicted exponent window, from about 0.18 to 1","key_machinery":"The central object is the dislocation gas Hamiltonian $H_D=-\\frac{Y}{8\\pi k_B T}\\sum_{i\\neq j} G(\\mathbf{r}_i-\\mathbf{r}_j):(\\mathbf{b}_i\\otimes\\mathbf{b}_j)+\\frac{\\epsilon_c+\\sigma}{k_B T}\\sum_i|\\mathbf{b}_i|^2$, where $G$ is the elastic Green function and the active stress $\\sigma=2\\eta\\dot\\epsilon$ is derived from the viscous stress of the convergent-extension flow around a T1 event. The constants $c_1$, $c_2$, and $c_3$ encode the statistical weight of one-, two-, and three-dislocation unbinding modes: $c_1$ is always finite, while $c_2$ is tied to T1 processes and $c_3$ to T2 processes, and in closed epithelia $c_3=0$. The renormalization-group flow for the softness $x$ and fugacity $y$, $dx/dl=c_2 y^2$ and $dy/dl=c_1 x y+c_3 y^2$, determines the divergence law for $\\xi_+$ and yields the exponent range in Eq. (5).","core_discovery":"The paper shows that the onset of collective cell migration in cell-resolved models of epithelial layers takes place via an activity-driven melting transition. Active forces, sourced by the convergent-extension flow of intercalating cells, shift the core energy of 5–7 dislocation pairs from $\\epsilon_c$ to $\\epsilon_c+\\sigma$, so the dislocation fugacity is $y=\\exp[-(\\epsilon_c+\\sigma)a^2/(k_B T)]$. Confluency reduces the degeneracy of dislocation-unbinding modes relative to point-particle crystals: in closed monolayers only T1 processes operate, giving $c_3=0$ and a Kosterlitz–Thouless-type flow, while open monolayers that permit T2 processes recover a finite $c_3$. The correlation length across the solid–hexatic boundary diverges as $\\log(\\xi_+/a)\\sim|\\sigma-\\sigma_c|^{-\\bar\\nu}$, with $\\bar\\nu$ between $\\frac{1}{2}\\left(1-\\frac{c_3}{\\sqrt{c_3^2+4c_1c_2}}\\right)$ and $1$. In numerical simulations of a multiphase-field model, closed systems give $\\bar\\nu=1.0\\pm0.1$ and open systems give $\\bar\\nu=0.18\\pm0.06$. When T1 processes are suppressed, the transition is replaced by a solid-to-solid transition with power-law divergence $\\xi\\sim|\\sigma-\\sigma_c|^{-\\nu}$, with $\\nu=c_1|x_0|$.","pith_inferences":["The exponent $\\bar\\nu$ measured in experiments could serve as a diagnostic for whether a tissue behaves as a closed monolayer (T1 only), an open monolayer (T1 plus T2), or an intermediate regime, offering a direct link between tissue turnover rates and critical behavior.","A decisive test of the equilibrium-like assumption would be to compare the temperature-like parameter extracted from the variance of cell displacements with the value inferred from fitting the dislocation fugacity: if the two disagree, the Boltzmann mapping used here does not hold even near criticality.","The predicted solid-to-solid transition when T1 is suppressed suggests that modulating cell intercalation—for instance via drugs that alter junctional tension—could switch a tissue between exponential and power-law correlation-length scaling, a testable pharmacological prediction.","Because the multiphase-field model allows T2 rates to be controlled externally, the same simulation framework could be used to map the full $\\bar\\nu$ versus T2-rate curve predicted by the RG analysis, providing a continuum check between the closed and open limits."],"forward_implications":["Increasing cell speed drives a two-step melting sequence from crystalline solid to hexatic liquid crystal to isotropic liquid, mediated first by unbinding of neutral 5–7 dislocation pairs and then by their splitting into isolated 5- and 7-fold disclinations.","In closed epithelial layers, where only T1 processes are available, the correlation-length exponent sits at the upper bound of the predicted range, $\\bar\\nu$ near 1, whereas open layers with T2 processes sit near the lower bound, $\\bar\\nu$ near 0.18, making the critical behavior tunable by controlling cell division and extrusion.","If T1 processes are suppressed while T2 processes remain, the transition is no longer of KTHNY type; instead the correlation length diverges as a power law, $\\xi\\sim|\\sigma-\\sigma_c|^{-\\nu}$ with $\\nu=c_1|x_0|$, a transition between two solid phases driven by persistent intrusion and extrusion events.","Extensile active stress can produce a negative effective core energy, rendering the crystalline solid unstable to cell intercalation, which is consistent with experimental observations on migrating MDCK monolayers.","The exponential divergence $\\log(\\xi_+/a)\\sim|\\sigma-\\sigma_c|^{-\\bar\\nu}$ provides a sharp, quantitative signature of the solid–hexatic boundary that can be sought in in vitro experiments."],"supporting_citations":[{"why":"Supplies the Kosterlitz–Thouless transition and the original renormalization-group treatment for the c3 = 0 case.","marker":"[14]"},{"why":"Provides Kosterlitz's calculation of the nu_bar = 1/2 exponent for the pure Kosterlitz–Thouless case.","marker":"[15]"},{"why":"Gives the dislocation-unbinding modes and the constants c1, c2, and c3 used in Equations (3) for two-dimensional triangular crystals.","marker":"[16]"},{"why":"Supplies the vector Coulomb gas description of two-dimensional melting and the same renormalization-group constants.","marker":"[17]"},{"why":"Establishes the correspondence between T1 rearrangements in epithelia and unbinding of 5–7 dislocations, the basis for attributing active stress to core energy renormalization.","marker":"[21]"},{"why":"Provides experimental evidence that contractile cells can form extensile monolayers, supporting the sign sigma < 0 used to connect activity with intercalation.","marker":"[27]"},{"why":"Supplies the multiphase-field model of confluent epithelia used in the simulations, including the active self-propulsion and the cell extrusion/intrusion implementation.","marker":"[43, 44]"}],"fun_headline_variants":["T1 and T2 defects tune epithelial melting exponents","Closed vs open tissues melt with different critical exponents","Active melting in epithelia has tunable universality class","Suppressing T1 swaps melting for sublimation in cell layers","Critical exponents for epithelial migration range 0.18 to 1"],"cache_read_input_tokens":22400,"weakest_assumption_plain":"The whole calculation rests on the premise that near the melting point the active, self-propelled cell layer behaves like a thermal equilibrium system, so that the standard equilibrium statistical mechanics of defects applies.","fun_headline_variants_meta":{"raw":{"variants":["T1 and T2 defects tune epithelial melting exponents","Closed vs open tissues melt with different critical exponents","Active melting in epithelia has tunable universality class","Suppressing T1 swaps melting for sublimation in cell layers","Critical exponents for epithelial migration range 0.18 to 1"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000622,"raw_usage":{"total_tokens":2959,"prompt_tokens":1100,"completion_tokens":1859,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":716,"completion_tokens_details":{"reasoning_tokens":1778}},"tokens_in":716,"tokens_out":1859,"duration_ms":11913,"temperature":1.0,"reasoning_tokens":1778,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T21:02:32.715489+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the density of unbound 5–7 dislocations and the positional correlation length while independently varying the self-propulsion speed and the rate of cell intrusion/extrusion, and check whether the fitted exponent falls in the predicted range and whether the divergence is exponential across the solid–hexatic boundary. More directly, test the equilibrium assumption by measuring the fluctuation–dissipation ratio near criticality: if the effective temperature extracted from that ratio does not match the value used in the Boltzmann fugacity, the equilibrium mapping fails.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the Kosterlitz–Thouless transition and the original renormalization-group treatment for the c3 = 0 case."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides Kosterlitz's calculation of the nu_bar = 1/2 exponent for the pure Kosterlitz–Thouless case."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the vector Coulomb gas description of two-dimensional melting and the same renormalization-group constants."},{"cited_title":"Krommydas, L","cited_arxiv_id":null,"evidence_quote":"Establishes the correspondence between T1 rearrangements in epithelia and unbinding of 5–7 dislocations, the basis for attributing active stress to core energy renormalization."},{"cited_title":"Balasubramaniam, A","cited_arxiv_id":null,"evidence_quote":"Provides experimental evidence that contractile cells can form extensile monolayers, supporting the sign sigma < 0 used to connect activity with intercalation."}],"review_version":1}