{"id":"5696c96c-4680-49f7-bbb4-e09e34ee6b02","arxiv_id":"2509.04087","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":8,"one_line_summary":"Ferritin diffusion in polymer crowder solutions follows a c*-normalized non-monotonic curve with a crossover near 2c*, attributed to depletion-induced intermediate-range order that bulk viscosity cannot explain.","lead":"Using a megahertz X-ray laser, researchers tracked how ferritin proteins move in solutions of sugar and polymer crowders. They found that polymer crowders create transient nanoscale order that changes protein diffusion in ways bulk viscosity alone cannot explain.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Dextran 500 at 10% and 35% w/w lacks the low-q D(q) signature required for IRO, directly contradicting the universal IRO-above-c* claim; the model-derived S(q) is not independently verified.","rationale":"The reader's weakest assumption already flagged the heavy reliance on two-Yukawa fits to D(q) and the dextran 500 exception. I agree that this is the soft spot, but I sharpen it to the explicit contradiction between the universal 'IRO above c*' statement and the SI's admission that dextran 500 at 10% and 35% w/w (both above c*) lacks the low-q D(q) slope change. This is not a minor detail: the abstract's first quantitative claim is universality, and the scaling collapse in Fig. 4C includes dextran 500. If the IRO interpretation is wrong for these two points, the two-Yukawa model is not universally applicable, and the derived S(q), H(q), K1, and lifetimes for those conditions are not trustworthy. The proposed test—extending to lower q or using contrast-matched SAXS—would directly resolve whether the absence is an artifact of q-range or a genuine counterexample. I do not think this warrants REJECT because the core observation that bulk viscosity alone cannot predict ferritin diffusion (the D_HS/D_SE deviation) is likely robust and supported by multiple crowder types. However, the universal IRO claim must be tempered or verified. Hence CONDITIONAL remains appropriate, with the condition being the successful outcome of the low-q/contrast-variation test for dextran 500.","tokens_in":21242,"tokens_out":11598,"duration_ms":102451,"concrete_test":"Re-analyze the raw XPCS data for dextran 500 at 10 and 35 %w/w using the full low-q range available from the 10 keV campaign (down to q ≈ 0.075 nm^-1) or, if necessary, perform a new SAXS experiment with contrast variation (e.g., D2O/sucrose matching) to isolate the ferritin-ferritin structure factor S(q). Check whether a low-q upturn or peak appears below the current q minimum. If no IRO signature is found, the claim that IRO occurs whenever c/c* > 1 must be revised; if it does appear, the original q-range was insufficient and the universality claim can be retained.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The central claim that IRO emerges whenever c/c* > 1 is contradicted by the paper's own SI ('Sticky hard sphere fits'): for dextran 500 at 10 and 35 %w/w, where c/c* = 2.0 and 7.0, the characteristic low-q slope change in D(q) is absent, and a purely attractive sticky-hard-sphere potential suffices. The SI dismisses this via 'limited q-range' or 'non-linear concentration dependence,' but neither is tested. Because the IRO interpretation is not based on a directly measured S(q)—the SI ('Limitations in S(q) determination') states that ferritin-crowder cross-correlations prevent unambiguous extraction of S_ferr-ferr(q)—the two-Yukawa fit (Eq. 2; SI Eq. 6) is the only evidence for IRO. If dextran 500 genuinely lacks IRO at these concentrations, then (i) the universality claim fails, and (ii) including dextran 500 in the c/c* scaling of D_HS/D_SE (Fig. 4C) may introduce a spurious 2c* crossover, since its derived S(q), H(q), and D0 are model-dependent. The self-diffusion D_HS is more robust (Table S2 vs S3 give similar D0), but the structural IRO claim—a headline result—is not.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports megahertz XPCS measurements of ferritin diffusion in sucrose, Ficoll, and three dextran molecular weights, extracting the q-dependent collective diffusion coefficient D(q). A two-Yukawa potential fitted to D(q) is used to obtain static structure factors S(q), hydrodynamic functions H(q), depletion strengths K1, complex lifetimes, and the microscopic self-diffusion coefficient D_HS. The central claims are that depletion-induced short-range attraction combined with long-range repulsion produces intermediate-range order (IRO) once the polymer overlap concentration c* is exceeded, and that normalizing crowder concentration by c* reveals a universal non-monotonic scaling of ferritin self-diffusion with a crossover near 2c*. The paper concludes that bulk viscosity alone cannot predict protein dynamics in polymer-crowded solutions.","tokens_in":21622,"tokens_out":3363,"duration_ms":35186,"significance":"If the conclusions hold, the work would provide a nanoscale dynamical counterpart to the well-known structural IRO phenomenology and would strengthen the case that polymer-specific interactions, not just excluded volume and bulk viscosity, control protein transport in crowded media. The experimental effort is substantial: MHz-XPCS at the European XFEL, careful control experiments showing that crowders alone give no XPCS signal, deposition of raw data at two DOIs, and explicit modeling of D(q) with competing interaction potentials. The paper is therefore potentially significant for soft-matter and biophysics audiences. However, the central quantitative results are model-derived rather than directly measured, and the paper's own SI contains statements that weaken the universality claim.","major_comments":[{"comment":"The universal-IRO claim is contradicted by the paper's own SI for dextran 500. The main text states that IRO emerges once c* is exceeded and the Discussion calls IRO 'a universal effect observed for both dextran and Ficoll.' Yet the SI reports that for dextran 500 at 10 %w/w and 35 %w/w, where c/c* = 2.0 and 7.0, the characteristic low-q slope change in D(q) is absent and a purely attractive sticky-hard-sphere potential describes the data. The SI attributes this to 'limited q-range' or 'non-linear concentration dependence,' but no test of either explanation is provided. Because the IRO interpretation is based on the fitted two-Yukawa model rather than on a directly measured S(q), this exception directly undermines the headline claim. Please either provide independent evidence for IRO in these cases or revise the universality claim.","section":"Results, Fig. 2E and SI 'Sticky hard sphere fits'"},{"comment":"The structural and dynamical quantities used as evidence are outputs of a two-Yukawa potential fitted to the same D(q) data, not independent measurements. Since D(q) = D0 H(q)/S(q) and both S(q) and H(q) are computed from the fitted potential, the K1 trends in Fig. 3D, the complex lifetimes in Fig. 3E, and the D_HS values used in Fig. 4A,C all reduce to properties of the fit. The SI explicitly states that ferritin-crowder cross-correlations prevent unambiguous isolation of S_ferr-ferr(q). Thus the model is the only evidence for IRO. Please validate at least one of these derived quantities by an independent route (e.g., contrast-matched SAXS, a model-free q-dependent feature, or a direct measurement of self-diffusion) before quantitative conclusions are drawn.","section":"Eq. (2) and SI 'Limitations in S(q) determination'"},{"comment":"The universal 2c* crossover in D_HS/D_SE is constructed using D_HS values that come from different interaction models for dextran 500 at 10 %w/w and 35 %w/w (two-Yukawa in Table S2 vs. sticky-hard-sphere in Table S3), while the paper also notes that these two concentrations lack the IRO signature. Including such points in the scaling plot without demonstrating that the crossover is robust to their exclusion or to a common model could produce a spurious universal trend. Please show the scaling plot with dextran 500 excluded or with a consistent model, and discuss how the model dependence of D_HS affects the crossover.","section":"Fig. 4C and Tables S2/S3"}],"minor_comments":[{"comment":"The text attributes the simulation predictions to 'Riest et al. [104]', but reference [104] is Fries et al. (2025) on chemically active droplets. Please correct the citation to the actual Riest and Nägele work.","section":"Discussion, reference [104]"},{"comment":"The units for D0(ccr) are given as nm2/µm, which appears to be a typo for nm2/µs. Also, the SI text mentions varying 'scl 1' but the table lists λ1; please unify the notation.","section":"Table S2 and SI 'Modelling of S(q) and H(q)'"},{"comment":"Since the fits use a stretched exponential with α = 0.9, D(q) = Γ(q)/q^2 is only an effective collective diffusion coefficient. The paper could state explicitly that deviations from α = 1 are small enough not to affect the reported D(q) trends, or quantify the systematic uncertainty.","section":"Eq. (1) and extraction of D(q)"}],"recommendation":"major_revision","confidential_remarks":"The manuscript's own SI is commendably transparent but it exposes a direct tension with the abstract's universality claim. The referee report focuses on the need to reconcile the dextran 500 exception and to validate at least one model-derived quantity independently. In addition, the incorrect citation of Riest et al. as [104] should be fixed. The paper is not fatally flawed; the raw XPCS data and the comparative model analysis are valuable, but the load-bearing IRO and scaling claims need substantial additional support or careful restatement."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"You should know this paper for the data, not for the model. The MHz-XPCS measurements of ferritin in sucrose, dextran (three molecular weights), and Ficoll are a real experimental effort, and the systematic comparison across crowder types and concentrations is new. The observation that the ratio of microscopic to Stokes-Einstein diffusion is non-monotonic in c/c*, with a crossover near 2c*, is interesting and likely robust. The control showing pure crowders produce no correlation signal is a good touch. If you work on crowding or XPCS, this is worth reading.\n\nThat said, the soft spots are real and they matter. The central structural claim — IRO above c* — is not backed by a directly measured S(q). The SI explicitly says that ferritin-crowder cross-correlations prevent unambiguous extraction of the ferritin-ferritin structure factor. Everything structural, including the IRO lifetime and the attractive strength K1, comes out of a two-Yukawa potential fit to the same D(q) used to define the effect. That is circular enough to weaken the quantitative claims.\n\nThe more concrete problem is dextran 500. At 10 and 35 %w/w, where c/c* is 2 and 7, the low-q slope change in D(q) is absent, and a purely attractive sticky-hard-sphere potential fits. That contradicts the universal claim that IRO appears whenever c > c*. The SI hand-waves this with 'limited q-range' or 'non-linear concentration dependence,' but neither is tested. The stress-test note is correct: including dextran 500 in the c/c* scaling plot may be loading the dice for a 2c* crossover. The authors should either resolve the dextran 500 behavior or soften the universality claim.\n\nAlso, the citation to [104] in the Discussion is wrong — it points to a paper about chemically active droplets, not the Riest simulation. And the analysis code is only available upon request, which will slow reproduction.\n\nThe paper is not fatally flawed. The raw data and the viscosity-decoupling observation are likely solid, and the issues are addressable: propagate errors through the model, re-analyze dextran 500 with a more honest justification, fix the citation, and deposit code. A serious referee should engage with this, not desk-reject it. I'd bring it to a reading group to discuss the fitting strategy, but I wouldn't treat the derived lifetimes or K1 values as hard numbers.","headline":"A genuinely useful XPCS dataset with systematic crowder comparison, but the headline IRO and 2c* claims lean on two-Yukawa fits to the same D(q) and have a known dextran 500 exception — still deserving of serious refereeing.","tokens_in":22310,"tokens_out":1903,"would_cite":true,"duration_ms":20761,"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":"Depletion-induced intermediate-range order, not bulk viscosity, governs nanoscale protein diffusion in polymer-crowded solutions once the overlap concentration is exceeded, with self-diffusion crossing over near 2c*.","keywords":["macromolecular crowding","depletion interaction","protein self-diffusion","intermediate-range order","X-ray photon correlation spectroscopy","polymer overlap concentration","ferritin","two-Yukawa potential"],"falsifier":"A neutron or contrast-matched SAXS measurement that renders the crowder invisible would show directly whether the low-q structure-factor peak (IRO) appears above c* in the same samples where D(q) changes slope; if no peak appears, the two-Yukawa interpretation fails. Independently, single-particle tracking of labeled ferritin across the 10–35 %w/w range should reproduce the non-monotonic ratio of microscopic to macroscopic self-diffusion with a turning point near 2c*, without relying on any hydrodynamic model.","tokens_in":21092,"feed_emoji":"🧬","tokens_out":17061,"duration_ms":149913,"temperature":0.7,"pith_summary":"Using coherent megahertz X-ray photon correlation spectroscopy to watch the iron-rich protein ferritin diffuse in sucrose, Ficoll, and three sizes of dextran, the paper sets out to show that polymeric crowders change protein dynamics through entropic depletion forces rather than through viscosity alone. Its central result is that once a polysaccharide crowder exceeds its overlap concentration c*—the point where polymer chains begin to overlap—depletion-driven short-range attractions compete with polymer-mediated long-range repulsions, generating transient intermediate-range order: clusters spanning two to three protein diameters that appear and vanish on microsecond-to-millisecond timescales, measurably altering the collective diffusion coefficient. Normalizing crowder concentration by c* collapses the measured self-diffusion across all polysaccharide crowders onto a common curve with a crossover near 2c*, below which a lower-viscosity depletion layer around each protein enhances mobility and above which rising bulk viscosity takes over. If correct, this means bulk viscosity is not a valid proxy for protein mobility in crowded environments, and models of cellular transport, association kinetics, and phase separation must include polymer-specific depletion interactions, crowder molecular weight, and overlap concentration.","feed_headline":"Proteins in polymer crowds change regime at 2× overlap concentration","feed_subtitle":"Above 2×c*, protein self-diffusion switches from depletion-enhanced to viscosity-limited motion.","key_machinery":"The load-bearing identity is D(q) = D0(ccr)·H(q)/S(q), which connects the measured collective diffusion coefficient to the static structure factor S(q) and the hydrodynamic function H(q). The argument is carried by a two-Yukawa potential—one short-range attractive Yukawa term plus one long-range repulsive term—fitted to the measured D(q) for each crowder and concentration; S(q) is obtained through the Ornstein-Zernike equation with a mean-spherical closure, and H(q) through the Beenakker-Mazur δγ-expansion. Everything downstream passes through that fit: the depletion attraction strength K1, the potential-well depth, the high-q limit of H(q) that yields the microscopic self-diffusion coeffici","core_discovery":"The paper's discovery is that the dynamical regime of a protein in a polymer crowder solution is set by the polymer overlap concentration c*, not by the absolute polymer concentration or the bulk viscosity. For ferritin in sucrose—a small-molecule crowder—the collective diffusion D(q) keeps the monotonic shape of a purely repulsive colloid at every concentration studied. In dextran and Ficoll solutions, once the crowder exceeds c*, the same D(q) develops a low-q modulation, which the authors attribute to intermediate-range order (IRO): short-range depletion attraction, from polymer configurational entropy lost where chains are excluded near the protein surface, balanced against a long-range","pith_inferences":["The c*-normalized scaling suggests a practical protocol for crowding studies: report polymer concentration in units of c* and treat 2c* as the regime boundary; this normalization may reconcile apparently contradictory literature reports on crowding effects.","Because the two-Yukawa fit is the single funnel through which S(q), H(q), K1, lifetimes, and D_HS all pass, the quantitative edifice inherits the potential's adequacy; an independent measurement of S(q) with cross-correlations suppressed (for example using contrast-matched crowders) would test the IRO interpretation directly.","If the scaling generalizes to other proteins, crowding near 2c* should reverse the common expectation that higher viscosity always slows binding: encounter rates could peak around the crossover, a testable prediction for reaction kinetics in dextran and Ficoll.","The paper's own sticky-hard-sphere fits show that a purely attractive potential describes dextran 500 data at 10 and 35 %w/w where the low-q IRO signature is absent, suggesting the two-Yukawa picture may need modification at extreme molecular weights—or that the accessible q-range hid the upturn."],"forward_implications":["Bulk viscosity cannot serve as a universal predictor of protein mobility in polymer-crowded solutions; the local depletion layer, polymer correlation length, and crowder molecular weight must enter models of crowding.","The crossover near 2c* provides a practical boundary for in vitro crowding experiments: below it, depletion effects enhance protein mobility; above it, viscosity-dominated slowing sets in.","Because complex lifetimes and attraction strength grow with crowder concentration and molecular weight, crowding can prolong protein-protein association times, partially offsetting the slowdown of diffusion-based encounter rates.","Depletion-induced IRO reduces hydrodynamic hindrance and raises sedimentation coefficients, implying that transport and phase-separation tendencies in crowded media depend on crowder identity in ways macroscopic rheology cannot capture.","The collapse of self-diffusion data onto a single c/c* curve across Ficoll and three dextrans indicates a common mechanism governed by polymer overlap rather than a crowder-specific effect."],"supporting_citations":[{"why":"Riest and Nägele's simulation of short-time dynamics under competing short-range attraction and long-range repulsion; it is the prediction the paper's IRO diffusion data validate.","marker":"[43]"},{"why":"Nägele's relation D(q)=D0·H(q)/S(q); it is the central equation from which S(q), H(q), and the self-diffusion coefficient are extracted.","marker":"[81]"},{"why":"Beenakker and Mazur's δγ-expansion for the hydrodynamic function; it converts the fitted interaction potentials into the H(q) used in all model D(q) calculations.","marker":"[82, 83]"},{"why":"Liu et al.'s two-Yukawa fluid model; it supplies the potential form used to fit D(q) in polysaccharide solutions and the cluster-formation framework.","marker":"[75]"},{"why":"Liu et al.'s characterization of intermediate-range order in lysozyme; it defines IRO as transient clusters and provides the structural interpretation of the low-q signal.","marker":"[39]"},{"why":"Shvets and Semenov's theory of effective interactions mediated by free polymer in semi-dilute solution; it grounds the long-range repulsion needed for the two-Yukawa picture.","marker":"[78]"},{"why":"Abkenar et al.'s dissociation-rate expression; it gives Eq. 3 used to estimate the lifetime of IRO-stabilized ferritin complexes.","marker":"[89]"},{"why":"Tuinier et al.'s depletion-layer model for sphere motion; it supplies Eq. 12 for the reduced microscopic viscosity inside the depletion layer.","marker":"[103]"},{"why":"Kohli and Mukhopadhyay's study of nanoparticle diffusion in semidilute polymer solutions; it documents the enhanced local diffusion the paper invokes to explain why measured self-diffusion exceeds bulk-viscosity predictions.","marker":"[46]"}],"fun_headline_variants":["Protein motion in polymer crowds flips at 2× overlap concentration","Depletion forces set protein diffusion crossover at 2c*","Polymer overlap concentration dictates protein mobility switch","At 2× overlap, protein diffusion changes from boosted to slowed","Crowder overlap marks shift from depletion boost to viscosity drag"],"cache_read_input_tokens":2688,"weakest_assumption_plain":"The results stand or fall on the assumption that a two-Yukawa model potential fitted to the measured diffusion curve correctly extracts the protein's structure factor, hydrodynamic function, and depletion-layer parameters—quantities the experiment cannot isolate directly, because ferritin-crowder cross-correlations contaminate the scattering signal and because the expected low-q order signature is missing in two of the dextran 500 samples.","fun_headline_variants_meta":{"raw":{"variants":["Protein motion in polymer crowds flips at 2× overlap concentration","Depletion forces set protein diffusion crossover at 2c*","Polymer overlap concentration dictates protein mobility switch","At 2× overlap, protein diffusion changes from boosted to slowed","Crowder overlap marks shift from depletion boost to viscosity drag"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000524,"raw_usage":{"total_tokens":2381,"prompt_tokens":766,"completion_tokens":1615,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":510,"completion_tokens_details":{"reasoning_tokens":1531}},"tokens_in":510,"tokens_out":1615,"duration_ms":11188,"temperature":1.0,"reasoning_tokens":1531,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-05T10:24:02.547195+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A neutron or contrast-matched SAXS measurement that renders the crowder invisible would show directly whether the low-q structure-factor peak (IRO) appears above c* in the same samples where D(q) changes slope; if no peak appears, the two-Yukawa interpretation fails. Independently, single-particle tracking of labeled ferritin across the 10–35 %w/w range should reproduce the non-monotonic ratio of microscopic to macroscopic self-diffusion with a turning point near 2c*, without relying on any hydrodynamic model.","supporting_citations":[{"cited_title":"Rep.272(5-6):215–372","cited_arxiv_id":null,"evidence_quote":"Nägele's relation D(q)=D0·H(q)/S(q); it is the central equation from which S(q), H(q), and the self-diffusion coefficient are extracted."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Shvets and Semenov's theory of effective interactions mediated by free polymer in semi-dilute solution; it grounds the long-range repulsion needed for the two-Yukawa picture."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Abkenar et al.'s dissociation-rate expression; it gives Eq. 3 used to estimate the lifetime of IRO-stabilized ferritin complexes."},{"cited_title":"Lett.75(6):929","cited_arxiv_id":null,"evidence_quote":"Tuinier et al.'s depletion-layer model for sphere motion; it supplies Eq. 12 for the reduced microscopic viscosity inside the depletion layer."}],"review_version":1}