{"id":"2e859df5-6526-4f78-89d7-c3a30cb3cf7c","arxiv_id":"2505.06153","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"Chemically active droplets tune the rate of bimolecular encounters, with an optimal acceleration when the droplet residence time matches the intra-droplet encounter time.","lead":"A simulation and mean-field modeling study shows that biomolecular condensates can either speed up or slow down the time two proteins need to meet, depending on droplet size, number, and how long proteins remain inside. The key result is that reaction kinetics are fastest when a protein's residence time in a droplet equals the time needed for two proteins to encounter each other inside that droplet.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The claimed optimum may rest on the slowest relaxation mode rather than the mean encounter time; the reported Te is not the mean first-passage time used for the homogeneous reference.","rationale":"The reader's weakest assumption concerns exponential waiting times and neglected correlations, but the reader also notes in the rationale that the reported mean encounter time is actually the slowest relaxation time. The present critique sharpens that second point: because the homogeneous reference T0 is a true mean first-passage time, the ratio Te/T0 mixes two different statistics, and the central optimum in Fig. 3g is defined in terms of the slowest mode. This is the most load-bearing concern because the headline result, that matching tau_res to tau_S,d maximizes the encounter rate, rests on the mean-field model's definition of Te. A concrete test comparing the slowest-mode time with the integral of the survival probability would settle whether the optimum is an artifact of the statistic chosen. The concern does not change the verdict: the paper is already CONDITIONAL, and this issue is a reason to keep that conditionality rather than to accept or reject outright. No code or data are provided, so the proposed recalculation is the appropriate next step.","tokens_in":19078,"tokens_out":5983,"duration_ms":67485,"concrete_test":"Recompute both statistics from the same mean-field equations and the same BD trajectories: (i) the slowest-mode time Te = -1/lambda_max, and (ii) the true mean first-passage time \\bar T = \\int_0^\\infty S(t) dt, truncated once S(t) reaches numerical zero, with S(t) the survival probability. Then regenerate Fig. 3g as \\bar T/T0 versus tau_res/tau_S,d, and repeat for the client systems in Fig. 4. If the minimum in \\bar T/T0 occurs at tau_res approximately tau_S,d and the acceleration regime survives, the central claim is supported; if the minimum shifts, disappears, or becomes monotonic, the claim must be revised.","verdict_should_be":"UNCHANGED","load_bearing_attack":"In the five-state model, p1(t) is a superposition of five exponentials, and the appendix defines Te as the largest characteristic time, i.e., Te = -1/lambda_max. The BD survival probability is likewise fitted to the long-time exponential decay (SI Fig. S2). This is not the mean first-passage time, which would be the integral \\bar T = \\int_0^\\infty (1-p1(t)) dt = \\sum_i w_i T_i. The homogeneous reference T0 = V/(8\\pi D\\sigma) is a Smoluchowski mean encounter time, so the ratio Te/T0 compares a slow-mode time with a mean time. The slowest mode can be dominated by rare pairs trapped in distinct droplets, and its amplitude can be small; the minimum in Fig. 3g at tau_res approximately tau_S,d may reflect the behavior of that rare mode rather than typical encounter kinetics. Since the central time-scale-matching claim is validated only through this Te definition, and the client BD comparison in Fig. 4b shows a monotonic decrease rather than a clear non-monotonic optimum, the quantitative claim that matching maximizes the encounter rate is not yet secured.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript combines Brownian dynamics (BD) simulations of phase-separating Lennard-Jones particles with a five-state mean-field model (Eqs. 1-5) to study bimolecular encounter kinetics in systems containing one or many droplets. The mean-field model is parameterized with transport coefficients and residence/entry times extracted from the same BD trajectories. For a single equilibrium droplet the model reproduces the BD encounter times and collapses onto Te/T0 = 1.14 D_c v/(D_d V); for multiple droplets it predicts Te/T0 ~ gamma N D_c v/(D_d V). In chemically active emulsions the chemical drive controls the residence time, and the mean-field model predicts that the encounter time is minimized when the residence time matches the intra-droplet Smoluchowski time. Client-protein BD simulations show qualitative agreement with the mean-field model, and the authors conclude that condensates can either accelerate or decelerate molecular encounters relative to a homogeneous system.","tokens_in":19340,"tokens_out":6418,"duration_ms":62475,"significance":"If the time-scale-matching principle survives scrutiny, it would provide a simple design rule for condensate-mediated reaction kinetics and a useful counterpoint to the common 'nanoreactor' picture. The work's strengths are the microscopic stochastic model, the explicit two-species client setup, and the honest bottom-up parameterization: the equilibrium comparison in Fig. 2b is a clean consistency check with no fitted parameters, and the client simulation series (Fig. 4b) is a nontrivial test of the mean-field approach. The significance is currently limited by the fact that the headline optimum is demonstrated only in the mean-field model under a definition of Te that is not the mean first-passage time, and the multi-droplet scaling used for biological estimates has no direct BD validation. These issues are fixable, but they are central to the claims.","major_comments":[{"comment":"The encounter time Te is defined as the largest characteristic time, Te = -1/lambda_max, and in BD it is extracted from the long-time exponential decay of the survival probability (SI Fig. S2). This is not the mean first-passage time \\bar{T} = \\int_0^\\infty S(t) dt that the homogeneous reference T0 = V/(8\\pi D\\sigma) represents. The ratio Te/T0 therefore compares a slow-mode relaxation time with a mean encounter time, and the minimum in Fig. 3g may be dominated by a rare slow mode with small amplitude. The paper should recompute the true mean first-passage time from the five-state model (the weighted sum of the mode times) and from BD (the average first encounter time) and verify that the matching optimum and the reported acceleration/deceleration ratios are preserved.","section":"Appendix 'Mean encounter time between Brownian particles'; Section II, Eqs. (1)-(5)"},{"comment":"The central claim that matching tau_res to tau_S,d maximizes the encounter rate is supported by mean-field calculations (Fig. 3g) but not directly by Brownian dynamics. In Fig. 4b, the client BD data show a monotonic decrease of Te/T0 with increasing Delta mu', with no clear nonmonotonic optimum, and the agreement with the mean-field model is only qualitative. Please provide a BD scan (for example, using the client drive) that resolves whether the nonmonotonic optimum exists in explicit particle simulations, or alternatively limit the claim to the mean-field model.","section":"Section VI and Fig. 4b"},{"comment":"The multi-droplet scaling Te/T0 = gamma N D_c v/(D_d V) is derived only from the mean-field model, with gamma fitted in the range 5-10, and no Brownian dynamics data at N > 1 are presented. This scaling is subsequently used to estimate biological slowdowns (for instance, a 15-fold increase for N = 6 x 10^3). A direct BD test at moderate N would be needed to support the multi-droplet deceleration prediction; otherwise the scaling should be presented explicitly as a mean-field extrapolation.","section":"Section III and Fig. 2d"}],"minor_comments":[{"comment":"The labels for panels (e) and (f) appear swapped: panel (e) plots tau_res and panel (f) plots J_A, while the caption assigns 'Net rate of A escape' to (e) and 'Residence time' to (f).","section":"Fig. 3 caption"},{"comment":"Calling Te the 'mean first encounter time' is misleading because it is the long-time decay time, not the average of the first-encounter distribution; please rename it or justify the definition in terms of the slowest mode.","section":"Section II"},{"comment":"The text contains 'Oswald ripening' where 'Ostwald ripening' is intended, and the Fig. 4 caption says 'For Delta mu = 18, droplets are unstable' where the context indicates Delta mu' = 18.","section":"Section III and Fig. 4 caption"},{"comment":"The model assumes exponential waiting times for droplet entry and exit, but the distributions of residence times and entry times are not reported; showing these distributions (alongside the survival probability in SI Fig. S2) would help justify the Markovian five-state description.","section":"Section II and SI"}],"recommendation":"major_revision","confidential_remarks":"The main risk is the mismatch between the reported Te (slowest relaxation mode) and the Smoluchowski mean first-passage time T0; if the authors recompute the true mean first-passage time and the optimum disappears, the headline would need substantial revision. The multi-droplet scaling is also thin because it rests on mean-field output only. I see no indication of misconduct; the issues are methodological and fixable within the manuscript's scope."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Main thing you should know: this paper is worth a look, but the central quantitative claim is not fully nailed. The authors build a five-state mean-field model that extends Smoluchowski kinetics to compartmentalized media, feed it parameters extracted from their own Brownian dynamics simulations, and show that encounter kinetics in active emulsions are optimized when the residence time in a droplet matches the Smoluchowski encounter time inside it. That time-scale-matching principle is genuinely new and physically plausible. The model is simple, the derivation is clear, and the single-droplet equilibrium comparison to BD (Fig. 2b) is clean, with no fitted parameters. The client simulations in Fig. 4b also show qualitative agreement without fitting.\n\nNow the soft spot. The stress-test note is right on the money. The appendix defines Te not as the mean first-passage time but as the largest characteristic time—the slowest relaxation mode of the survival probability. The BD values are likewise extracted from the long-time exponential tail. The homogeneous reference T0 is a true Smoluchowski mean encounter time. So the dimensionless ratio Te/T0 mixes a slow-mode time with a mean time. If the slow mode is dominated by a rare population of pairs trapped in distinct droplets, the minimum in Fig. 3g at tau_res ≈ tau_S,d may not describe typical encounters. The authors are explicit about this definition, so it is not a hidden flaw, but the central claim that matching maximizes the encounter rate rests on it. The client BD comparison is monotonic rather than showing a clear non-monotonic optimum, which is consistent with the concern.\n\nA second, smaller point: the multi-droplet scaling Te/T0 = gamma N D_c v/(D_d V) has gamma fitted, and there is no BD validation for N > 1. The parameters are also extracted from the same simulations the model is tested against, so the agreement is a consistency check rather than an external prediction. No code or data are provided, which makes it harder to assess.\n\nWho is this for? People working on biomolecular condensates and reaction kinetics. The idea that chemical activity can switch condensates between accelerators and decelerators of reactions is important for the field. The paper deserves a serious referee, but the referee should ask for a clearer statistical definition of Te—ideally the true mean first-passage time or at least an analysis of the amplitude of the slow mode. My own verdict would be conditional: the qualitative design principle is likely right, but the quantitative claim needs another pass.","headline":"A useful mean-field model for encounter kinetics in active emulsions with a qualitatively plausible time-scale-matching optimum, but the reported 'mean encounter time' is actually the slowest relaxation mode, so the quantitative claim needs another pass.","tokens_in":19803,"tokens_out":2262,"would_cite":true,"duration_ms":22460,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"Chemically active droplets can either speed up or slow down molecular encounters, and the fastest encounter rate is reached when a particle's residence time in a droplet matches the Smoluchowski encounter time inside that droplet.","keywords":["biomolecular condensates","liquid-liquid phase separation","encounter kinetics","chemically active droplets","Brownian dynamics","mean-field model","residence time matching","Smoluchowski theory"],"falsifier":"Measure the residence time distribution of client proteins in a condensate system and the first-encounter time between partners while sweeping the chemical drive. If the encounter rate shows no maximum when τres crosses the droplet Smoluchowski time, or if the maximum appears at a very different ratio, the central claim fails. Concretely, a single-particle tracking experiment with a tunable ATP-driven modification should show a non-monotonic mean encounter time versus Δμ.","tokens_in":18888,"feed_emoji":"💧","tokens_out":5890,"duration_ms":54745,"temperature":0.7,"pith_summary":"Condensates formed by liquid-liquid phase separation are often described as nanoreactors that concentrate reactants and speed up reactions, but they can also trap molecules in separate compartments and slow encounters down. This paper argues that in chemically active droplets the outcome is controlled by a time-scale match: the rate of bimolecular encounters is maximized when the residence time of a particle in a droplet equals the Smoluchowski encounter time inside the droplet. The claim is supported by Brownian dynamics simulations of phase-separating particles with stochastic chemical interconversion, together with a five-state mean-field model whose rates are taken from the simulations. If correct, the result turns condensates into tunable regulators: cells could repress or activate specific protein encounters by adjusting the chemical drive that sets residence times.","feed_headline":"Droplet reactions peak when stay time equals meeting time","feed_subtitle":"Chemically active condensates can speed or slow molecular encounters; the optimum sits at a time-scale match.","key_machinery":"The load-bearing object is a five-state mean-field model of a two-particle search in a medium divided into N droplets of volume v and a continuous phase. The five states are: the particles have met; both are in distinct droplets; both are in the same droplet; one is in a droplet and the other in the continuous phase; both are in the continuous phase. Transitions between states are governed by inverse times: Smoluchowski encounter times τS,d = v/(8πD_dσ) inside a droplet and τS,c = (V−Nv)/(8πD_cσ) in the continuous phase, a residence time τres in a droplet, and an entry time τent. Solving the master equations yields the mean first encounter time Te as the longest relaxation time. The simulations supply τres, τent, diffusivities, droplet volumes, and droplet numbers, and the model then predicts Te without fitted parameters, reproducing the simulation data and exposing the optimum at τres ≈ τS,d.","core_discovery":"The paper's central claim is that active chemical reactions do not merely set droplet size and stability; they directly control molecular transport through the droplet interface, and this transport controls reaction kinetics. In active emulsions driven by a chemostat, the residence time τres decreases as the chemical drive Δμ increases, generating particle fluxes at droplet surfaces. The key result is that, for fixed droplet geometry, the mean first encounter time Te is minimized when τres ≈ τS,d = v/(8πD_dσ), the Smoluchowski time for two particles to meet inside a droplet of volume v. When residence is much longer, particles waste time trapped in separate droplets; when much shorter, droplets do not concentrate partners long enough. The authors show that condensates can therefore either accelerate or decelerate encounters relative to a homogeneous reference, and they demonstrate the optimum in Brownian dynamics simulations with client proteins whose drive is varied independently of the scaffold.","pith_inferences":["If the residence-time-matching rule is generic, synthetic condensate reactors could be engineered by choosing a chemical drive that sets τres equal to the predicted τS,d for the target pair.","A testable extension would be to measure, in a single experimental condensate system, whether varying the chemical drive produces the predicted non-monotonic encounter-time curve with a minimum at τres ≈ τS,d.","The same framework could be applied to multi-reactant networks, where selective acceleration or repression of individual pair encounters might shape reaction pathways without changing protein concentrations.","Because the mean-field model neglects correlations between the two searching particles, the optimum may be sharper or shifted in crowded condensates where cooperative transport matters."],"forward_implications":["A condensate can act as either an activator or a repressor of a given bimolecular reaction, depending on the ratio τres/τS,d.","Because the chemical drive sets τres, post-translational-type modifications coupled to a free-energy reservoir could tune encounter rates in cells without changing droplet size or number.","In multi-droplet systems at equilibrium, the encounter time grows approximately linearly with the number of droplets, so non-specific condensates can act as traps that slow partner search.","Even with identical droplet geometry, active systems can have different encounter kinetics, so structure alone does not determine reaction speed in active emulsions.","The five-state model generalizes Smoluchowski diffusion-limited kinetics to compartmentalized media and could be used to design condensate-based reactors."],"supporting_citations":[{"why":"Supplies the Smoluchowski encounter time formula τS = V/(8πDσ) for homogeneous media that the compartmentalized model adapts.","marker":"[23]"},{"why":"Provides the microscopic stochastic model of chemically active droplets and the simulation method that the paper extends to encounter kinetics.","marker":"[47]"},{"why":"Establishes the active droplet behavior and enzymatic-reaction control that motivate the residence-time analysis.","marker":"[48]"},{"why":"Supplies the framework of active chemical reactions coupled to phase separation that underlies the chemostat and droplet size selection.","marker":"[38]"},{"why":"Defines the purely repulsive interaction potential used for the homogeneous reference and for non-attracting species.","marker":"[50]"},{"why":"Provides the random telegraph process formalism used for the stochastic A⇌B interconversion reactions.","marker":"[54]"},{"why":"Provides the biological multi-droplet scenario (RNA granules in C. elegans) used to estimate whether condensates accelerate or slow encounters.","marker":"[34]"}],"fun_headline_variants":["Droplet reactions peak when stay equals meet","Active droplets tune encounter rates both ways","Reaction optimum: droplet stay time matches meet time","Chemically active droplets control reaction speed","Droplet residence sets encounter kinetics"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The five-state mean-field model treats droplet entry, exit, and encounters as independent exponential waiting times with constant rates, and it ignores correlations between the two searching particles; if real residence times are non-exponential or the two partners' stays are correlated, the predicted optimum at τres ≈ τS,d could shift or disappear.","fun_headline_variants_meta":{"raw":{"variants":["Droplet reactions peak when stay equals meet","Active droplets tune encounter rates both ways","Reaction optimum: droplet stay time matches meet time","Chemically active droplets control reaction speed","Droplet residence sets encounter kinetics"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000197,"raw_usage":{"total_tokens":1343,"prompt_tokens":903,"completion_tokens":440,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":519,"completion_tokens_details":{"reasoning_tokens":374}},"tokens_in":519,"tokens_out":440,"duration_ms":5424,"temperature":1.0,"reasoning_tokens":374,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T22:47:22.497008+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the residence time distribution of client proteins in a condensate system and the first-encounter time between partners while sweeping the chemical drive. If the encounter rate shows no maximum when τres crosses the droplet Smoluchowski time, or if the maximum appears at a very different ratio, the central claim fails. Concretely, a single-particle tracking experiment with a tunable ATP-driven modification should show a non-monotonic mean encounter time versus Δμ.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the Smoluchowski encounter time formula τS = V/(8πDσ) for homogeneous media that the compartmentalized model adapts."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the microscopic stochastic model of chemically active droplets and the simulation method that the paper extends to encounter kinetics."},{"cited_title":"S¨ oding, D","cited_arxiv_id":null,"evidence_quote":"Establishes the active droplet behavior and enzymatic-reaction control that motivate the residence-time analysis."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the framework of active chemical reactions coupled to phase separation that underlies the chemostat and droplet size selection."},{"cited_title":"Berthin, J","cited_arxiv_id":null,"evidence_quote":"Defines the purely repulsive interaction potential used for the homogeneous reference and for non-attracting species."},{"cited_title":"Mazzocca, A","cited_arxiv_id":null,"evidence_quote":"Provides the random telegraph process formalism used for the stochastic A⇌B interconversion reactions."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the biological multi-droplet scenario (RNA granules in C. elegans) used to estimate whether condensates accelerate or slow encounters."}],"review_version":1}