{"id":"3657800d-6fec-4806-8599-91b22a714469","arxiv_id":"2509.06087","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":6,"one_line_summary":"In overlapping-circle confinement, the ratio of circle separation to radius controls whether MDCK epithelial sheets rotate, oscillate with periodic direction reversals, or move disorderly; a self-propelled particle model reproduces the sequence.","lead":"Changing the overlap of two circular cell-culture islands switches a sheet of kidney cells from rotating as a vortex to sloshing back and forth, and then to moving chaotically. A simple computer model of self-propelled cells reproduces the switch and suggests it is governed by a balance of two cell-cell interaction rules.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Mechanistic claim that LA/CIL balance drives the Δ/R transition is underdetermined: the model is tuned via μ_c=0.5, omits traction/adhesion/heterogeneity, and its quantitative mismatches (threshold 1.17 vs 1.33, period ~1000 vs ~360 min) prevent it from uniquely supporting the proposed mechanism.","rationale":"The reader's weakest assumption is exactly the mechanism claim: it rests on the polarity equation with LA and CIL only, is supported by tuning μ_c to 0.5, and is not tested experimentally; the authors acknowledge omitted traction and mechanical heterogeneity. I agree with that identification. My stress-test adds that the model's quantitative mismatches (transition ratio 1.17 vs 1.33, period 1000 vs 360 min, differing order-parameter definitions, and the internal 1.33/1.44 inconsistency) make the simulation's support for the proposed mechanism weaker than the phrase 'reproduced' suggests. However, these problems do not overturn the paper's primary empirical finding that Δ/R correlates with a vortex-to-oscillatory-to-disordered sequence; they only prevent acceptance of the mechanistic interpretation. Since the reader's verdict is already CONDITIONAL, my assessment does not change the verdict. The proposed test—experimental CIL perturbation—would directly resolve whether the mechanism claim is necessary or merely a possible interpretation.","tokens_in":13264,"tokens_out":6071,"duration_ms":69729,"concrete_test":"Repeat the doublet confinement experiments (R=150 μm, Δ/R from 0 to 1.97) with MDCK cells in which CIL is suppressed—e.g., by N-cadherin knockdown or by treatment with a Rho/ROCK pathway inhibitor—and measure the vortex order parameter and oscillation period versus Δ/R. If the transition at Δ/R≈1.33 and the period peak persist under CIL suppression, then the LA/CIL balance is not necessary for the observed dynamics and the mechanistic claim fails; if the transition shifts or disappears, the proposed mechanism is supported.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The abstract and conclusion claim not just that Δ/R controls the transitions but that 'an appropriate balance between LA and CIL is critical for stabilizing vortex pairs with velocity reversals.' This mechanistic claim rests entirely on the SPP model, Eq. (3), where polarity evolves under only local alignment and CIL, with μ_c=0.5 selected because it reproduces the experimental transition. That selection is not an independent test. The quantitative agreement is loose: the simulated vortex-order transition occurs at Δ/R≈1.17 versus the experimental ≈1.33 (Results and Fig. 4), the simulated peak period is ~1000 min (600τ at τ=1.67 min) versus the measured ~6 h and is called 'comparable,' and the experimental order parameter is computed in a 20-μm boundary strip while the simulated one uses the whole domain. The paper itself is inconsistent about the critical ratio, giving 1.33 in the Results and 1.44 in the Discussion. The model omits cell-substrate traction, adhesion mechanics, and mechanical heterogeneity, an omission the Discussion explicitly acknowledges, and the Discussion leaves open whether active polar fluid or active vertex models could capture the same transition. Because no experimental perturbation of CIL or alignment is performed, the data cannot distinguish the proposed LA/CIL mechanism from the traction-driven density-wave mechanism previously proposed for rectangular confinement (refs 41–43). Thus the empirical claim that Δ/R is a control parameter is reasonably supported, but the mechanistic part of the central claim is not established.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript studies MDCK epithelial monolayers confined in a 'doublet' geometry formed by two partially overlapping circles of radius R. Varying the center separation ratio Δ/R from 0 to ~2 changes the collective motion: a stable co-rotating vortex for Δ/R<1.33, oscillatory flow along the long axis with a ~6 h period, and ultimately disordered motion. The authors introduce a self-propelled particle model with local alignment (LA) and contact inhibition of locomotion (CIL), show that with moderate CIL (μ_c=0.5) the simulated sequence matches qualitatively, and sweep CIL strength to argue that an LA/CIL balance is critical. The abstract and conclusion claim that Δ/R is a minimal control parameter and that the LA/CIL balance stabilizes vortex pairs with velocity reversals.","tokens_in":13471,"tokens_out":6947,"duration_ms":73965,"significance":"If established, the paper would provide a simple geometric handle on active turbulence in epithelial sheets and a bridge between circular and rectangular confinement results. The experimental design is clean and the PIV analysis standard. The model is simple and transparent, and the μ_c sweep is a useful exploration. However, the mechanistic conclusion is currently supported only by the model with a calibrated coupling constant, not by experimental perturbation; the quantitative agreement is loose; and the 'minimal control parameter' claim is not tested at multiple R for the order parameter. With those caveats, the empirical transition sequence is a valuable contribution and the model generates falsifiable predictions about CIL strength, although these predictions are not experimentally tested.","major_comments":[{"comment":"The mechanistic claim in the abstract and conclusion ('an appropriate balance between LA and CIL is critical') rests entirely on the SPP model. In Eq. (3), polarity evolution contains only LA and CIL; μ_c=0.5 is selected because it reproduces the observed transition (Results, Fig. 5). The μ_c sweep shows that the output changes with this free parameter, but this is not an independent test: no experimental perturbation of CIL or alignment is performed. The Discussion acknowledges omitted traction, adhesion, and mechanical heterogeneity and leaves open active polar-fluid or active-vertex alternatives. Therefore the data do not distinguish the proposed LA/CIL mechanism from, e.g., the traction-driven density-wave mechanism of refs. 41–43. Either add CIL/alignment perturbation experiments or substantially weaken the causal language.","section":"Eq. (3) and Fig. 5"},{"comment":"The claim that the model 'successfully reproduced' the experimental transitions is undermined by quantitative mismatches. The simulated vortex-order-parameter transition occurs at Δ/R≈1.17, whereas the experimental transition is Δ/R≈1.33 (Fig. 4(b) vs Fig. 3(b)). The simulated peak period is ~1000 min (600τ), compared to the measured ~6 h (360 min), and describing this as 'comparable' overstates the agreement. In addition, the experimental order parameter is computed in a 20-μm boundary strip while the simulation uses the whole domain, so the matching values (~0.3) are not directly comparable. These discrepancies need to be addressed or the language needs to be qualified to 'qualitative agreement.'","section":"Results, 'To complement...' and Figs. 3–4"},{"comment":"The critical aspect ratio is quoted inconsistently. The Results and Fig. 3(b) state the experimental vortex-order transition at Δ/R≈1.33; the Discussion states 'approximately Δ/R=1.44 in both the experiment and the simulations' and later 'critical value Δ/R=1.44'; the simulation order-parameter transition is Δ/R≈1.17 while the period peak is at Δ/R≈1.44. The authors should define a single estimation procedure for the transition and period maximum and report consistent values with uncertainties.","section":"Discussion"},{"comment":"The phrase 'minimal control parameter governing transitions' is too strong given the data. The order-parameter measurements are shown only for R=150 μm; the period measurements for R=100 and 125 μm are reported but no order-parameter transition for these radii. Because vortex stability in circular confinement depends on R (Fig. 1(i)), the claim that Δ/R alone determines the transition requires testing at multiple R for the order parameter or an explicit scaling argument.","section":"Results, doublet geometry"}],"minor_comments":[{"comment":"The statement 'Δ/R = cos Ψ' cannot hold for Δ/R>1 (the experiments reach 1.97). Either the angle is misdefined or a factor of 2 is missing. This should be corrected.","section":"Results, doublet boundary definition"},{"comment":"The text says 'a circular boundary with a diameter of R=150 μm' while elsewhere R is the radius; the l_v/2≈150 μm comparison indicates R=150 μm is the radius. Please correct.","section":"Fig. 1(e)"},{"comment":"The harmonic-potential boundary term introduced in the Methods is not quantified; its strength is a free parameter that could affect the transition. Please specify it or state that it is a numerical regularization.","section":"Numerical simulation"},{"comment":"The ordinate in Fig. 2(e) is denoted k_t^{-1}; define this as the oscillation period and state how peak-to-peak intervals are converted.","section":"Fig. 2"},{"comment":"Several typographical issues: 'ordered vortex rotation was observed' (Introduction), 'IW AKI' for IWAKI, inconsistent reference bracket spacing. Also the definition of the neighbor set A_i via Voronoi tessellation with a 3σ_r cutoff should be clarified (e.g., are Voronoi neighbors within cutoff only?).","section":"General"}],"recommendation":"major_revision","confidential_remarks":"The experimental core is sound and potentially of broad interest, but the mechanistic overreach and the quantitative mismatches are likely to be sticking points. If the authors can either add a perturbation experiment (e.g., modulating CIL or alignment) or fully recast the conclusion as a model-motivated hypothesis, the paper could be publishable. As is, I recommend major revision."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Dear X,\n\nThe thing to know: the empirical core is real and the aspect-ratio control is a useful single knob; the mechanistic story about LA/CIL balance is not backed by the experiments. I would send this to review, but I'd want a serious revision that separates the data from the model inference.\n\nThe paper does something new: it transfers the doublet geometry from the group's bacterial vortex work to MDCK monolayers, sweeps Delta/R continuously, and shows a vortex-to-oscillation-to-disorder sequence in both experiment and simulation. The vortex order parameter dropping from ~0.3 to near 0 as Delta/R passes ~1.33 is a clean observation, and the ~6 h oscillation period with a peak near the transition is interesting. Computing the order parameter in a boundary strip is sensible, and the simulations reproduce the qualitative sequence with a simple LA+CIL model. That is real progress beyond independent vortex and rectangular-oscillation studies.\n\nThe soft spots are all around the mechanism claim. The paper concludes that a moderate balance between LA and CIL is critical, but the only support is a parameter sweep retaining mu_c=0.5 because it fits. There is no experimental perturbation of CIL or alignment, so the data cannot distinguish the proposed mechanism from the traction-driven density-wave picture of refs 41–43. The model also omits substrate traction and mechanical heterogeneity, which the authors acknowledge. That is a real limitation, and the central conclusion overstates what is shown. The quantitative mismatches are minor but symptomatic: the simulated transition at Delta/R~1.17 vs the experimental 1.33, the simulated period ~1000 min vs ~6 h called 'comparable', and the paper itself gives 1.33 in Results and 1.44 in Discussion for the same curve. None of this kills the empirical claim, but it means the model is a proof of concept rather than a test.\n\nWho should read it: anyone working on geometric control of collective cell migration, and the active matter community generally. It deserves peer review, but the authors should be pushed to tone down the mechanism claim and either add a perturbation experiment or reframe the model as illustrative.\n\nMy recommendation: send to review, with the expectation of major revision.","headline":"Empirical aspect-ratio control is solid and useful; the LA/CIL mechanism claim is model-selected, not experimentally tested.","tokens_in":14169,"tokens_out":2058,"would_cite":true,"duration_ms":24564,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"A single geometric ratio switches confined epithelial cells between vortex, oscillation, and disorder.","keywords":["collective cell migration","epithelial monolayer","confinement aspect ratio","vortex rotation","oscillatory motion","contact inhibition of locomotion","self-propelled particle model","micropatterned confinement"],"falsifier":"Knock down or inhibit contact inhibition of locomotion in MDCK cells confined to doublet circles, then remeasure the vortex order parameter, oscillation period, and disorder threshold across Δ/R. The model predicts that weakening CIL stabilizes vortex pairs and shifts the transition, while strengthening CIL destroys vortices at all aspect ratios; if the experimental transition curve is unchanged, the LA/CIL balance is not the controlling mechanism.","tokens_in":12997,"feed_emoji":"🌀","tokens_out":7318,"duration_ms":74041,"temperature":0.7,"pith_summary":"The paper sets out to show that one number—the aspect ratio of a doublet-shaped confinement, defined as the distance between the centers of two overlapping circles divided by their radius—controls which collective motion an epithelial sheet chooses. In experiments on MDCK cells, small aspect ratios give a stable rotating vortex, intermediate values give back-and-forth oscillatory flow with a period of about six hours, and large aspect ratios give disordered motion. The authors reproduce the same three-state sequence in a self-propelled particle model in which cells align locally and turn away from neighbors on contact, and they identify the balance between those two rules as the cell-level mechanism. If the claim holds, tissue engineers and active-matter physicists get a minimal geometric knob for programming collective behavior without changing any biochemical parameter.","feed_headline":"Aspect ratio alone flips cells from vortex to oscillation to disorder","feed_subtitle":"Changing only the spacing of two overlapping circular wells moves cell sheets from rotating to oscillating to disorganized.","key_machinery":"The argument is carried by two objects. First, the doublet circular boundary: two overlapping circles of fixed radius R whose center separation Δ tunes the confinement aspect ratio Δ/R while preserving the circular boundary conditions known to stabilize vortices; changing Δ/R changes the anisotropy and the collision geometry of the two would-be vortices. Second, the polarity dynamics of the self-propelled particle model (Eq. 3), in which each particle's direction turns under local alignment (LA), which pulls it toward the velocity direction of neighbors, and contact inhibition of locomotion (CIL), which pushes it away from the relative position of neighbors; the strengths μ_a and μ_c set the","core_discovery":"The paper identifies the confinement aspect ratio Δ/R of a doublet circular boundary—two equal circles of radius R with centers separated by Δ—as a minimal control parameter for epithelial collective motion. For Δ/R below about 1.33, MDCK cells form a stable co-rotating vortex; above it, the population oscillates, reversing its migration direction along the long axis with a period of about 6 h that peaks near Δ/R≈1.33–1.44 before falling; at larger aspect ratios motion becomes disordered. A self-propelled particle model with local alignment and moderate contact inhibition of locomotion reproduces this sequence, while weak CIL stabilizes counter-rotating vortex pairs and strong CIL destroys o","pith_inferences":["If the geometric rule is generic, the same vortex-oscillation-disorder sequence should appear in other confined active fluids (bacterial suspensions, active nematics, cell extracts) placed in doublet boundaries; testing this would separate geometry-driven physics from MDCK-specific biology.","The paper's mechanism could be wrong even if the transition is real: rectangular-confinement oscillations have been explained by traction-driven density waves, and the doublet geometry may simply be another route to the same physics, with LA/CIL being a phenomenological stand-in.","A direct test would be to pharmacologically or genetically weaken contact inhibition of locomotion in MDCK cells and remeasure the transition curve; the model predicts the boundary and period maximum should shift systematically.","The maximum of the oscillation period near Δ/R≈1.33–1.44 resembles a critical slowing-down signature, suggesting that a hidden ordering transition (possibly vortex-pairing) sits at that aspect ratio; measuring correlation times and fluctuation amplitudes could reveal critical exponents."],"forward_implications":["Collective mode in a confined epithelial sheet can be predicted from geometry alone: measure Δ/R and the state (vortex, oscillation, disorder) follows, without needing cell-type-specific biochemical parameters.","The oscillation period grows linearly with Δ up to a maximum near Δ/R≈1.44 because a cell traveling at constant speed needs time proportional to Δ to cross the doublet; beyond that the period drops as order breaks down.","Moderate contact inhibition of locomotion is required for oscillatory dynamics; too little CIL locks the system into vortex pairs and too much CIL produces disorder, so perturbations of CIL should shift the transition.","Doublet-shaped micropatterns offer a design rule for tissue engineering: by carving the overlap of two circles, one can program rotation, rhythmic pumping, or disorganized migration in a cultured epithelial layer.","The results connect epithelial oscillations to vortex-pairing transitions known in bacterial vortices, where two counter-rotating vortices stabilize when Δ/R exceeds √2."],"supporting_citations":[{"why":"Establishes that circular confinement stabilizes vortex-like collective modes and defines the characteristic length scale used to set the circle radius.","marker":"[16]"},{"why":"Shows emergence and persistence of collective migration on small circular micropatterns, supporting the circular-boundary vortex baseline.","marker":"[20]"},{"why":"Reports spatiotemporal oscillation in confined epithelial motion, providing precedent for velocity reversals in circular confinement.","marker":"[21]"},{"why":"Documents wavelike collective migration modes in rectangular confinement and gives the ~8 h oscillation period and size-independence that the doublet results are compared with.","marker":"[41]"},{"why":"Shows sustained oscillations of epithelial cell sheets, the main experimental precedent for oscillatory motion under confinement.","marker":"[42]"},{"why":"Provides a theoretical treatment of spontaneous and induced oscillations in confined epithelia, framing the possible mechanisms behind the observed reversals.","marker":"[43]"},{"why":"Supplies the self-propelled particle framework from which the doublet-confinement simulation model is built.","marker":"[44]"},{"why":"Introduces the local-alignment and contact-inhibition-of-locomotion polarity dynamics (Eq. 3) that generate vortex and oscillatory states in the simulations.","marker":"[45]"},{"why":"Shows that counter-rotating bacterial vortices stabilize when the doublet separation exceeds about √2 times the radius, the comparison used for the MDCK vortex-pairing behavior.","marker":"[33]"}],"fun_headline_variants":["Change ring spacing to flip cell vortex to oscillation","Single parameter: vortex to oscillation by ring overlap","Cell motion shifts from swirl to crawl as rings separate","Aspect ratio triggers vortex, oscillation, disorder"],"cache_read_input_tokens":2688,"weakest_assumption_plain":"The explanation assumes that cell behavior is governed by just two local rules—aligning with neighbors and turning away on contact—with the strength of the second rule chosen to match the data; if traction, adhesion, or cell deformability actually drive the switching, the geometric transition could persist while the proposed cell-level mechanism fails.","fun_headline_variants_meta":{"raw":{"variants":["Change ring spacing to flip cell vortex to oscillation","Single parameter: vortex to oscillation by ring overlap","Cell motion shifts from swirl to crawl as rings separate","Aspect ratio triggers vortex, oscillation, disorder"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00023,"raw_usage":{"total_tokens":1319,"prompt_tokens":747,"completion_tokens":572,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":491,"completion_tokens_details":{"reasoning_tokens":512}},"tokens_in":491,"tokens_out":572,"duration_ms":6780,"temperature":1.0,"reasoning_tokens":512,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-05T04:30:17.752319+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Knock down or inhibit contact inhibition of locomotion in MDCK cells confined to doublet circles, then remeasure the vortex order parameter, oscillation period, and disorder threshold across Δ/R. The model predicts that weakening CIL stabilizes vortex pairs and shifts the transition, while strengthening CIL destroys vortices at all aspect ratios; if the experimental transition curve is unchanged, the LA/CIL balance is not the controlling mechanism.","supporting_citations":[{"cited_title":"G., Duclos, G","cited_arxiv_id":null,"evidence_quote":"Establishes that circular confinement stabilizes vortex-like collective modes and defines the characteristic length scale used to set the circle radius."},{"cited_title":"J., Th¨ uroff, F., Piera Alberola, A., Frey, E","cited_arxiv_id":null,"evidence_quote":"Shows emergence and persistence of collective migration on small circular micropatterns, supporting the circular-boundary vortex baseline."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Reports spatiotemporal oscillation in confined epithelial motion, providing precedent for velocity reversals in circular confinement."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Documents wavelike collective migration modes in rectangular confinement and gives the ~8 h oscillation period and size-independence that the doublet results are compared with."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Shows sustained oscillations of epithelial cell sheets, the main experimental precedent for oscillatory motion under confinement."},{"cited_title":"P., Pruitt, B","cited_arxiv_id":null,"evidence_quote":"Provides a theoretical treatment of spontaneous and induced oscillations in confined epithelia, framing the possible mechanisms behind the observed reversals."},{"cited_title":"& Jacob, E","cited_arxiv_id":null,"evidence_quote":"Supplies the self-propelled particle framework from which the doublet-confinement simulation model is built."},{"cited_title":"& Feng, X.-Q","cited_arxiv_id":null,"evidence_quote":"Introduces the local-alignment and contact-inhibition-of-locomotion polarity dynamics (Eq. 3) that generate vortex and oscillatory states in the simulations."},{"cited_title":"Proceedings of the National Academy of Sciences118,e2107461118 (2021)","cited_arxiv_id":null,"evidence_quote":"Shows that counter-rotating bacterial vortices stabilize when the doublet separation exceeds about √2 times the radius, the comparison used for the MDCK vortex-pairing behavior."}],"review_version":1}