{"id":"45f1fca6-584a-409c-99e4-662a343ac3b1","arxiv_id":"2511.18077","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"Motile bacteria regulate dextran–PEG phase separation through coupled activity and interfacial wetting, suppressing coarsening in droplet regimes but accelerating it when dextran is the minority phase.","lead":"Bacteria swimming in a mixture of two liquids can stop droplets from merging by making them spin, or—when one liquid is scarce—glue droplets into chains and branched clusters. The paper shows this happens only when swimming and the bacteria's preference for one liquid act together.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Simulation wetting and activity parameters are uncalibrated; claimed mechanisms may be an artifact of tuned γ1, γ2, f_act.","rationale":"The reader's weakest_assumption correctly identifies that the mechanistic conclusions rest on tuned simulation parameters (γ1, γ2, f_act) whose fidelity to the real P. aeruginosa system is unvalidated. My stress-test concurs: the two central mechanisms—rotation-induced hydrodynamic repulsion and wetting-mediated capillary attraction—are both demonstrated only in simulations with these hand-set values, and no calibration against the measured contact angle or swimming speed is shown. The missing Supplemental Material further prevents assessment of the model's details. Because the experimental morphology sequence is real but the proposed physical mechanisms are not independently verified, the CONDITIONAL verdict remains appropriate. I do not see a more load-bearing concern that would move the verdict to ACCEPT or REJECT; the concern is about quantitative fidelity, not internal inconsistency.","tokens_in":11304,"tokens_out":5602,"duration_ms":61815,"concrete_test":"Perform a parameter sweep in the FPD model over (γ1, γ2) consistent with the measured contact angle θ≈50° (using the Young equation in the model) and over f_act consistent with the measured swimming speed in the ATPS. Check whether the morphology sequence (spinning droplets → chains → branched clusters) and the two-sphere repulsion/attraction persist across the physically allowed range. Additionally, measure the contact angle in the simulation for the chosen γ1, γ2 and verify it equals ≈50°; if it does not, recalibrate and re-run the morphology sequence.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim—that activity–wetting coupling controls LLPS morphology and coarsening—is established primarily through FPD simulations. The wetting affinities γ1=−4 and γ2=2 are set 'motivated by the experimentally observed amphiphilicity' but are not derived from the measured contact angle θ≈50°; no mapping from θ to these coefficients is provided, and the Supplemental Material (ref. [54]) is a placeholder. Similarly, the force-dipole magnitude f_act is not calibrated to the swimming speed of P. aeruginosa in the DEX–PEG mixture. The two key mechanistic conclusions—(i) low-Re hydrodynamic repulsion between self-spinning droplets (Fig. 3d) and (ii) wetting-mediated capillary attraction between bacteria (Fig. 4c,d)—are demonstrated only in simulations with these hand-set parameters. If the real interfacial stress balance differs (e.g., the actual contact angle or active stress is different), the rotation-repulsion and capillary-attraction mechanisms may be simulation artifacts. Additionally, the abstract claims a biofilm-associated protein acts as a 'wetting glue' promoting bacterial clustering, but this result is absent from the main text and appears only in the placeholder Supplemental Material, so a key piece of evidence is unavailable.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper combines confocal experiments on Pseudomonas aeruginosa in a dextran–polyethylene glycol aqueous two-phase system with fluid-particle-dynamics (FPD) simulations. It reports a morphology sequence (self-spinning droplets, elongated droplet chains, branched capillary-like clusters, and strongly deformed droplets) controlled by the DEX volume fraction and bacterial volume fraction. The central mechanistic proposal is that bacterial activity and interfacial wetting act together: activity suppresses coarsening in the droplet regime via rotation-induced hydrodynamic repulsion, and accelerates coarsening in the low-DEX regime via wetting-mediated capillary attraction. The paper also reports a dual kinetic role for activity and proposes a general active-wetting coupling framework.","tokens_in":11651,"tokens_out":7123,"duration_ms":68485,"significance":"If the proposed mechanism is correct, the paper establishes a genuinely new control principle for liquid–liquid phase separation in active binary fluids: motility and wetting, not either effect alone, set the morphology and coarsening kinetics. The direct experimental observation of the morphology sequence and of the opposite effect of activity on coarsening in the two regimes is valuable and will likely stimulate follow-up work. However, the mechanistic conclusions are carried by FPD simulations whose wetting coefficients and activity force are set ad hoc, and a key biological claim (the \"wetting glue\" protein) is confined to a placeholder Supplemental Material. Quantitative kinetic curves are shown without error bars or replicate information. The work is therefore a promising but not yet fully supported contribution.","major_comments":[{"comment":"The two central mechanisms are established with the wetting coefficients γ1=−4, γ2=2 and activity f_act that are chosen \"motivated by the experimentally observed amphiphilicity\" with no mapping to the measured contact angle θ≈50° and no calibration of f_act to the swimming speed of P. aeruginosa. Since the morphology sequence is then reproduced with these same values, the agreement in Fig. 2 does not by itself validate the mechanism. Please provide a derivation of γ1, γ2 from the measured contact angle, or at minimum a sensitivity analysis over γ1, γ2, f_act; otherwise the rotation-repulsion and capillary-attraction mechanisms remain post-hoc.","section":"Hybrid particle–field simulations; Fig. 2"},{"comment":"The experimental evidence for the dual kinetic role is presented as single curves without error bars, replicate number, or measurement definitions for R and A. More importantly, the \"non-motile\" control is not described; if inactivation is achieved by heat or chemical treatment, the cell surface (and hence wetting affinity) may change, confounding activity with altered interfacial properties. The method of producing non-motile bacteria and the statistics must be reported, and ideally a flagellar/chemotaxis mutant should be used.","section":"Fig. 3(a), Fig. 4(a), Fig. A2"},{"comment":"The abstract and summary claim that a biofilm-associated protein acts as a \"wetting glue\" promoting bacterial clustering even in dilute suspensions. This result is not presented in the main text or End Matter; it is relegated to the placeholder Supplemental Material [54]. As submitted, this is an unverifiable core claim and should either be brought into the main text or removed from the abstract.","section":"Abstract and Ref. [54]"},{"comment":"The effective low-Reynolds-number repulsion between two rotating droplets is demonstrated only by a simulation described as \"two 3D rollers\" with details in the missing SM. Because Stokes-flow interactions between rotating bodies are sensitive to boundary conditions and to the presence of a deformable interface, it is crucial to report the domain size, boundary conditions, and droplet deformation in this test. The experimental time-lapse (Fig. 3e) shows separation, but alternative explanations (Marangoni flow, droplet deformation, or confinement) are not ruled out.","section":"Self-spinning droplets and hydrodynamic repulsion; Fig. 3(d)"},{"comment":"The comparison between experiment and simulation is exclusively visual/qualitative. No quantitative morphology metrics (e.g., droplet size distribution, chain length, cluster fractal dimension) or a phase diagram in (ϕdex, ϕbac) are provided. The claim that the mechanism is \"robust\" would be better supported by a quantitative mapping between experimental ϕdex/ϕbac and simulation ψ̄, and by showing that the sequence persists for a range of f_act and γ values.","section":"Fig. 2, Fig. 4"}],"minor_comments":[{"comment":"Typo: \"passive ATTS\" should presumably read \"passive ATPS.\"","section":"Introduction"},{"comment":"The axis label \"τt/ s\" and the caption's \"τt/ e\" are inconsistent; it should likely be t/τ_e.","section":"Fig. 3(a)"},{"comment":"The condition \"γ1, γ2 = 0\" should be written \"γ1 = γ2 = 0\" for clarity.","section":"Fig. 4(c)"},{"comment":"The Supplemental Material URL contains \"xxx\" and is a placeholder; it must be completed before final submission.","section":"Ref. [54]"},{"comment":"The caption says the DEX volume fraction \"increases from left to right\" but the panel labels may be ambiguous; clarify.","section":"End Matter, Fig. A2"}],"recommendation":"major_revision","confidential_remarks":"The stress-test concern about uncalibrated γ and f_act is valid in my reading. The authors should be asked for a parameter-sensitivity study and the full experimental methods/statistics before acceptance. The abstract's biofilm-protein claim should be either substantiated in the main text or removed."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nQuick take: the experiments deserve attention; the simulations do not yet carry the mechanistic weight the narrative puts on them. The paper finds that P. aeruginosa in a DEX-PEG ATPS drives a clean morphological sequence — self-spinning droplets, droplet chains, branched capillary clusters — as the DEX volume fraction drops below the bacterial fraction. That is new, and the observation that activity suppresses coarsening in one regime and accelerates it in another is a genuinely interesting handle on active phase separation. The confocal evidence, including non-motile controls and a PEG-minority control, is direct and convincing.\n\nThe FPD simulations reproduce the morphologies qualitatively and offer a microscopic story: rotating droplets hydrodynamically repel each other at low Reynolds number, and wetting-mediated chemical potential gradients generate capillary attraction between bacteria in the low-DEX regime. These are plausible mechanisms, and the two-particle simulations illustrate them clearly.\n\nThe soft spots are real but manageable. The wetting coefficients γ1=-4, γ2=2 and activity force f_act are set by hand, motivated by a measured contact angle that is not quantitatively mapped to these parameters, and the supplemental that would explain the mapping is a placeholder. So the mechanisms are not validated by the experiments; they are consistent with the experiments. That distinction matters. Also, the abstract announces a biofilm-associated protein acting as a 'wetting glue,' but that result is absent from the main text and supplement, so as presented it is an unsupported claim. Kinetics curves in Fig 3(a) and 4(a) would benefit from error bars.\n\nNone of this sinks the paper. The experimental morphology sequence and the dual kinetic effect are the main results, and they stand on their own. The simulation is a helpful interpretive scaffold, not a proof. The authors are candid about the discrepancies with their own model, which I find encouraging.\n\nMy recommendation: send it to peer review, but require the supplemental, a sensitivity analysis for the wetting and activity parameters, error bars or raw data for the kinetic plots, and a revision of the abstract to only claim what is actually shown. This is a serious contribution that needs tightening, not a desk reject.","headline":"Solid experimental morphology sequence with a plausible but under-calibrated simulation mechanism; deserves peer review with required revisions.","tokens_in":12060,"tokens_out":2748,"would_cite":true,"duration_ms":29797,"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":"This paper claims that bacterial motility and interfacial wetting act together as a single control mechanism for liquid–liquid phase separation in a dextran–polyethylene glycol mixture, giving activity a dual kinetic role that suppresses or","keywords":["active matter","liquid-liquid phase separation","aqueous two-phase system","Pseudomonas aeruginosa","wetting","hydrodynamic interactions","coarsening","droplet morphology"],"falsifier":"In an active suspension with dextran as the majority phase, track the center-to-center distance between two nearby self-spinning droplets: the mechanism predicts a systematic, monotonic separation without coalescence at Reynolds number around 10^-3, while a non-motile control should coalesce; failure to observe such separation would refute the hydrodynamic-repulsion explanation.","tokens_in":11230,"feed_emoji":"🦠","tokens_out":7208,"duration_ms":66492,"temperature":0.7,"pith_summary":"This paper argues that in a phase-separating aqueous two-phase system of dextran and polyethylene glycol, motile Pseudomonas aeruginosa act not as passive inclusions but as active agents whose swimming and wetting behavior jointly determine the pattern of phase-separated domains. The central claim is that activity converts self-propulsion into interfacial stresses, producing a sequence of morphologies—self-spinning droplets, elongated droplet chains, branched capillary-like clusters, and deformed droplets—as the dextran volume fraction is lowered. The paper further claims that activity has a dual kinetic role: hydrodynamic repulsion from droplet rotation suppresses coarsening in the droplet regime, while wetting-mediated capillary attraction accelerates coarsening when dextran is the minority phase. A sympathetic reader cares because this identifies activity–wetting coupling, rather than activity or wetting alone, as a physical control mechanism for liquid–liquid phase separation, suggesting practical routes to stabilize or destabilize emulsions and to promote bacterial aggregation.","feed_headline":"Bacteria can slow droplet growth or speed it up","feed_subtitle":"Swimming microbes turn a dextran-PEG mixture into spinning droplets, chains, or capillary clusters by phase ratio.","key_machinery":"The central object is a coarse-grained pusher-type swimmer: each bacterium is a head–tail rod with a force dipole of magnitude f_act, embedded in a binary fluid described by a Ginzburg–Landau free energy with order parameter ψ (DEX-rich vs PEG-rich). Wetting enters through affinity coefficients γ1 and γ2 on the head and tail, set to γ1 = −4 and γ2 = 2 to match P. aeruginosa's amphiphilicity with slight DEX preference. The fluid dynamics couples the order parameter to the Navier–Stokes equation via the fluid particle dynamics method. This machinery converts bacterial propulsion into interfacial stresses: in the droplet regime it produces coherent internal rotation; at low DEX fraction it prod","core_discovery":"The paper claims that the interplay between bacterial activity and interfacial wetting—not either alone—regulates liquid–liquid phase separation in a dextran–polyethylene glycol mixture. As the dextran fraction is lowered, experiments and simulations show a reproducible sequence: self-spinning droplets, elongated droplet chains, branched capillary-like clusters, and fingerlike deformed droplets. Activity suppresses coarsening in the droplet regime because bacteria inside droplets drive coherent rotation, and two rotating droplets undergo a low-Reynolds-number hydrodynamic repulsion that blocks coalescence. When dextran is the minority phase, activity accelerates coarsening: wetting affinity","pith_inferences":["This suggests a design rule for switchable emulsions: if bacterial swimming can be toggled by oxygen or nutrient availability, the same chemical system could be driven between slow and fast coarsening regimes.","The tuned wetting parameters imply a falsifiable prediction: altering bacterial surface chemistry should shift the morphology sequence and the sign of activity's effect on coarsening in a predictable way.","The two-bacteria simulations suggest a generic mechanism: any preferentially wetting small particle—not just a swimmer—could accelerate aggregation in a minority-phase-scarce regime, with activity amplifying the effect."],"forward_implications":["Motile bacteria can act as an active stabilizer: the same strain both delays macroscopic stratification in the droplet regime and accelerates domain growth in the capillary-cluster regime, so a single control parameter—the phase-volume ratio—selects the direction of coarsening.","The morphology of an all-aqueous emulsion can be programmed by adjusting the ratio of minority-phase volume to bacterial volume, yielding droplets, chains, or branched clusters without changing chemistry.","Wetting-mediated capillary attraction offers a physical route to bacterial aggregation from dilute suspension, potentially relevant to early biofilm formation.","The rotation-induced hydrodynamic repulsion at low Reynolds number extends the known colloidal wetting-and-coarsening picture to active swimmers, implying that hydrodynamic interactions cannot be neglected in active ATPS models."],"fun_headline_variants":["Bacteria slow or speed droplet ripening via wetting","Swimming bacteria spin droplets, then control coarsening","Wetting and motion give bacteria dual control of droplets","Bacteria's wetting trick flips droplet growth into fast or slow"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The entire mechanism rests on simulations where the wetting affinities (γ1 = −4, γ2 = 2) and the activity force were chosen to reproduce the observed morphology rather than independently measured; if that tuned interfacial stress balance is not quantitatively faithful to real P. aeruginosa, the proposed rotation-repulsion and capillary-attraction mechanisms could be artifacts.","fun_headline_variants_meta":{"raw":{"variants":["Bacteria slow or speed droplet ripening via wetting","Swimming bacteria spin droplets, then control coarsening","Wetting and motion give bacteria dual control of droplets","Bacteria's wetting trick flips droplet growth into fast or slow"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000348,"raw_usage":{"total_tokens":1728,"prompt_tokens":719,"completion_tokens":1009,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":463,"completion_tokens_details":{"reasoning_tokens":940}},"tokens_in":463,"tokens_out":1009,"duration_ms":10766,"temperature":1.0,"reasoning_tokens":940,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-03T20:47:29.351758+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"In an active suspension with dextran as the majority phase, track the center-to-center distance between two nearby self-spinning droplets: the mechanism predicts a systematic, monotonic separation without coalescence at Reynolds number around 10^-3, while a non-motile control should coalesce; failure to observe such separation would refute the hydrodynamic-repulsion explanation.","supporting_citations":[],"review_version":1}