{"id":"e59a345d-528c-4f85-b63e-ad659ae202df","arxiv_id":"1909.02722","paper_version":3,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"Stokes-derived friction and diffusion tensors in anisotropic Langevin dynamics give solvent-free membrane models realistic undulation kinetics, matching continuum predictions for free and wall-adjacent membranes.","lead":"This paper adds a realistic water-like friction and flow coupling to fast 'solvent-free' computer models of cell membranes, using solutions of the Stokes flow equations to build the missing hydrodynamic interactions. The method reproduces the slow rippling motions of membranes, both free-standing and near a wall, which matters for simulating cellular processes such as budding and signaling.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The Stokes-derived D tensor assumes an incompressible membrane (v_parallel=0), but the simulated membrane is compressible; the agreement with Seifert may be coincidental and needs a compressible-boundary test.","rationale":"The reader's verdict is CONDITIONAL based on several observations: partial circularity, fitted b, sensitivity to alpha, missing code/data, and a suspected factor error in Eq. (31). I agree that these warrant conditional acceptance, but the single most load-bearing physical concern is the internal inconsistency between the incompressible-membrane boundary condition used to derive the hydrodynamic tensors and the compressible membrane model actually simulated. The reader's weakest_assumption mentioned this incompressibility issue alongside the normal-only/pairwise-additive assumption; I focus on the incompressibility part because the pairwise additive superposition is actually exact for the flat, linearized problem, whereas the mismatch in boundary conditions is a genuine inconsistency that could affect the dispersion relations. The agreement with the Seifert continuum model, which explicitly includes density fluctuations, is strong evidence that the framework works, but only if the incompressible-BC derivation does not accidentally produce the correct mobility for the compressible case. The proposed test, replacing v_parallel = 0 with the compressible continuity condition and comparing D_ij, directly resolves whether the concern lands. Additionally, the factor-of-four discrepancy in the far-field comparison (Eq. (31) vs. the stated Oseen tensor) hints that the relationship between the derived D tensor and standard membrane hydrodynamics is not fully understood, further supporting the need for this check. Since these issues can be settled by additional analysis and do not invalidate the demonstrated fast-mode agreement, the appropriate verdict remains CONDITIONAL (unchanged from the reader).","tokens_in":20169,"tokens_out":21678,"duration_ms":244330,"concrete_test":"Independently re-derive the single-membrane response in Section IV using the compressible boundary condition v_parallel = w_parallel, with w_parallel determined by the membrane density field (continuity equation), and compute the resulting D_ij(r) from Eq. (28). If the corrected D_ij differs from the incompressible-BC result by more than ~10% for r ≲ 5a (the range that dominates nearest-neighbor couplings), implement it in the Section VI A simulations and check whether the slow dispersion branch still matches Eq. (34) without refitting b.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The load-bearing step is the translation from Stokes boundary conditions to the diffusion tensor. In Section IV, the single-membrane solution explicitly imposes the incompressible-membrane condition v_parallel = 0 at z=0 (the paragraph after Eq. (15): 'Assuming the membrane to also follow the continuity condition of an incompressible fluid, we have iq w̃_parallel(q) = 0'). This condition is used to fix A1(q) and obtain the stress/velocity responses (Eqs. 19 and 20), which are then integrated in Eq. (28) to produce D_ij. The particle-based membrane used in Section VI, however, has a finite area compressibility modulus K_area = 0.270 N/m and supports density fluctuations; the reference continuum model (Eq. 34) explicitly includes the density field rho_q and a slow 'slipping' mode driven by density relaxation. Thus the simulated membrane does not satisfy v_parallel = 0, and the D_ij used in the Langevin equation is the response of a more constrained system than the one simulated. The agreement with the Seifert model (which is compressible) in Figs. 4 and 5 is therefore not a direct validation: it may hinge on the fitted inter-leaflet friction b (chosen to match) and on the insensitivity of the fast mode to the boundary condition. If the correct compressible boundary condition changes the normal-normal mobility at the probed wavelengths, the predicted dispersion relations would shift away from Seifert.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper introduces a framework for coupling solvent-free particle-based membrane models to solvent hydrodynamics. The authors propose anisotropic Langevin dynamics with pairwise friction/diffusion tensors whose out-of-plane components are obtained from solutions of the Stokes equations under localized Gaussian velocity or stress boundary conditions on the membrane. Closed-form and semi-analytical results are derived for a single planar membrane, a pair of parallel planar membranes (also modeling a membrane near a wall), and a spherical vesicle. The tensors are then used in Langevin dynamics simulations of a previously developed particle-based membrane model, and the dispersion relations of undulation modes are compared with Seifert-type continuum predictions for free-standing membranes and membranes near a wall. The fast hydrodynamic mode agrees well with the continuum model, the equilibrium undulation spectrum is correct, and the near-wall results reproduce the expected wall-distance dependence. The slow slipping mode is matched by choosing the inter-leaflet friction coefficient b as a fitting parameter.","tokens_in":20491,"tokens_out":6261,"duration_ms":71467,"significance":"If the framework holds, it is a practically valuable contribution to multiscale membrane simulation: it offers a way to reintroduce long-range hydrodynamic interactions into solvent-free particle-based models without the cost of explicit solvent or grid-based solvers. The derivation from Stokes equations is transparent, and the agreement of the fast hydrodynamic mode with the Seifert continuum model is a nontrivial validation, since that mode is not tuned by any fitted parameter. The paper also demonstrates correct equilibrium sampling (the undulation power spectrum follows Eq. (33a)) and shows systematic improvement with increased hydrodynamic cutoff radius. These strengths make the method useful for the target community. The main weakness is the mismatch between the incompressible-membrane boundary condition used in deriving the diffusion tensors and the compressible nature of the simulated membrane and of the reference continuum model; this gap needs to be addressed before the central claim is fully supported.","major_comments":[{"comment":"The diffusion tensors used in the Langevin equation are derived under the incompressible membrane boundary condition iq w_parallel(q) = 0 at z = 0 (paragraph after Eq. (15)). However, the simulated membrane has a finite area compressibility modulus K_area = 0.270 N/m (Table I), and the reference continuum model (Eq. (34)) explicitly includes a density field rho_q and a slow slipping mode; the simulated membrane can therefore sustain in-plane velocity and density fluctuations. The agreement with Eq. (34) in Figs. 4 and 5 is thus not a direct validation of the derived mobility for the system being simulated; it may rely on the fitted inter-leaflet friction b and on the insensitivity of the fast mode to the in-plane boundary condition. Please address this by (i) deriving the normal-normal mobility under a compressible boundary condition and showing that the difference is negligible at the probed wavevectors, or (ii) demonstrating numerically that the in-plane membrane velocity in the simulated model is small enough that the incompressible condition is a good approximation for the relevant modes. Without such a check, the claim that the dispersion relations follow from first-principle hydrodynamics is not fully established.","section":"Sec. IV (after Eq. (15)) and Sec. VI A (Eq. (34), Table I)"},{"comment":"The inter-leaflet friction coefficient b is chosen as b = 10^6 N s/m^3 to give a good match between the theoretical frequencies and simulation results, and the text acknowledges that b mostly affects the slipping mode. Consequently, the filled-symbol data in Fig. 4(a,b) constitute a fit of the slow mode rather than a parameter-free prediction. The conclusion in Sec. VII states that the framework predicts 'realistic timescales' relying solely on properties such as bending modulus and solvent viscosity, 'removing the need for ad hoc corrections after the fact.' This statement is too strong for the slow mode. Please either provide an independent estimate of b (e.g., from the model's inter-leaflet friction or from density relaxation measurements) or explicitly scope the parameter-free claim to the fast hydrodynamic mode.","section":"Sec. VI A, paragraph beginning 'In using Eq. (34)...'"},{"comment":"For the near-wall case, the authors fit triple exponentials to the undulation autocorrelation and compare the dominant mode with Eq. (36), but the two additional fitted timescales are not validated against any continuum model. Since the membrane is compressible and the two leaflets experience different hydrodynamic environments, these modes likely arise from density and slipping dynamics, but no reference description is provided. Please either supply a continuum counterpart for these modes (for instance, an extension of Eq. (34) with wall boundary conditions) or state explicitly that the additional slow modes are phenomenological and not yet validated. This is important because the presence of a third slow timescale is presented as a qualitative result in the text.","section":"Sec. VI C and Fig. 5"}],"minor_comments":[{"comment":"The paper does not state whether the divergence terms ∇_j · D_ij in Eq. (2) (given in Eq. (5)) were computed and included in the reported simulations. For flat membranes they vanish, but for the near-wall simulations, small membrane curvature may make them nonzero. Please clarify whether these terms were included or neglected.","section":"Secs. III and VI"},{"comment":"The dispersion relations obtained from biexponential fits are presented without error bars or any measure of fit uncertainty. Given that the slow mode may have a small amplitude at some wavevectors, please report confidence intervals or at least the number of independent samples used for the fits.","section":"Sec. VI A, Fig. 4"},{"comment":"The notation 'lim_{rij→∞, rij ≫ rc}' is ambiguous. Please define r_c and state the precise limiting procedure used to obtain the far-field approximation D⊥_ij ≈ kT/(4πηr_ij).","section":"Eq. (31)"},{"comment":"Several references contain formatting errors, e.g., Ref. [81] 'H IROMI Y AMAKAWA', Ref. [87] 'T ransform. Appl. Handb.', and Ref. [91] 'GNU Scientific Library Reference, Technical Report July'. Please unify and correct the reference list.","section":"References"},{"comment":"The statement 'We believe this to be the only viable means to reliably investigate complex, membrane-involved biological processes close to their native time-scales' is too categorical for the evidence presented. Please soften it to reflect that the approach is one promising route.","section":"Sec. VII"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is a solid methods paper with a clear derivation and a credible fast-mode validation. The main concern is the incompressible-membrane boundary condition used to derive D_ij versus the compressible simulated membrane; I would like to see either a dedicated test of that assumption or a more careful framing of the validation. The fitted b for the slow mode is disclosed, but the conclusion overstates the parameter-free nature of the predictions; this should be fixed. I believe the issues are addressable within revision and do not warrant rejection."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The paper is worth a serious look. It introduces a practical way to put long-range hydrodynamics into solvent-free particle-based membrane models without explicit solvent, by deriving normal-normal diffusion tensors from Stokes-flow solutions for three geometries and plugging them into anisotropic Langevin dynamics. The validation is partial but the fast undulation mode genuinely matches Seifert's continuum result, and the equilibrium spectrum is right. That is a real result.\n\nWhat is new: the closed-form self mobilities (Eq. 29), the parallel-membrane and vesicle tensor calculations, and the demonstration that the scheme reproduces the correct hydrodynamic mode for free and near-wall membranes. The paper is clearly written about the derivation, and the assumptions (only normal-normal hydrodynamic coupling, pairwise additivity) are stated up front.\n\nSoft spots, in order of importance. First, the Stokes derivation assumes an incompressible membrane (v_parallel = 0 at the boundary), but the simulated membrane has finite area compressibility and supports density fluctuations. The reference continuum model is compressible. The fast mode may be insensitive to this, but the paper doesn't show that; the agreement could be partly coincidental. A test with a compressible boundary condition would settle it. Second, the inter-leaflet friction b is fitted to match the slow mode, and the slow mode also depends on the hydrodynamic width alpha. The authors note this, and the fast mode is robust to it, but it does weaken the claim that everything is first-principles. Third, no code or data is provided. For a computational methods paper, that's a meaningful barrier. Fourth, the far-field comparison in Eq. (31) says the result is half the Oseen zz component, but the factor looks off by a factor of four relative to the stated text. Minor, but should be fixed.\n\nOverall, the central idea holds up. The incompressibility worry is the one that could bite, and it needs either a numerical test or a theoretical argument. The paper deserves a serious referee; I'd ask the authors to address the compressible case and release the code/data. This is for people building coarse-grained membrane models for long-timescale processes; they'll want to know if the scheme works beyond the planar patch.","headline":"A useful, well-derived hydrodynamics scheme for solvent-free membranes, with a real but addressable compressibility caveat.","tokens_in":20957,"tokens_out":2736,"would_cite":true,"duration_ms":29849,"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":"The paper claims that anisotropic Langevin dynamics with Stokes-derived friction and diffusion tensors gives solvent-free particle-based membrane models the same large-scale undulation kinetics as continuum hydrodynamic theory.","keywords":["solvent-free membrane models","hydrodynamic interactions","anisotropic Langevin dynamics","Stokes equations","membrane undulations","dispersion relations","coarse-grained simulation","pairwise diffusion tensors"],"falsifier":"Repeat the same membrane model with explicit solvent or a grid-based hydrodynamic solver and compare the undulation dispersion relations, especially at short wavelengths and small particle separations; a systematic mismatch there would show where the normal-only pairwise tensor misses in-plane or many-body coupling. A second check is to use a much softer membrane, since the derivation's incompressible-boundary assumption should break down more visibly at high q.","tokens_in":19936,"feed_emoji":"💧","tokens_out":11939,"duration_ms":123863,"temperature":0.7,"pith_summary":"The paper tackles a known weakness of solvent-free coarse-grained membrane simulations: they can reproduce equilibrium membrane shapes but typically have unrealistic kinetics because the surrounding fluid is absent. It proposes to keep the model solvent-free and instead encode the fluid into anisotropic Langevin dynamics, with pairwise friction and diffusion tensors computed from Stokes-flow solutions for idealized membrane geometries. The claim is that this reproduces the wavelength-dependent relaxation rates (dispersion relations) of membrane undulations predicted by continuum elastic-membrane theory, for free-standing patches and for patches near a solid wall. If that is right, particle-based models gain the hydrodynamic kinetics needed for large-scale, millisecond-timescale studies of processes such as membrane remodelling, without the cost of explicit solvent or grid-based flow solvers.","feed_headline":"Stokes-flow tensors restore hydrodynamics to solvent-free membranes","feed_subtitle":"Pairwise friction and diffusion tensors from Stokes equations reproduce continuum undulation speeds in particle-based membrane models.","key_machinery":"The load-bearing object is the anisotropic diffusion tensor of Eq. (4) inside the overdamped Langevin update of Eq. (2). Its normal-only pair part means solvent-mediated forces between distinct particles act along membrane normals, which keeps the tensor assembly cheap and makes the divergence terms vanish for flat membranes. The out-of-plane response is computed from Stokes equations with a Gaussian boundary condition of width $\\alpha$ (Eq. (9)); Hankel transforms turn that boundary condition into stress or velocity response fields (Eqs. (19) and (20)), and numerical integration over hexagonal-lattice particle patches yields the friction and diffusion coefficients of Eq. (28). The derivation enforces a no-slip, incompressible membrane at the boundary, which separates these kernels from plain point-force (Oseen) tensors and leads to a far-field pair diffusion of roughly $kT/(4\\pi\\eta r_{ij})$—half the $zz$ component of the Oseen tensor. In-plane mobility is appended separately via the standard diffusion formula for inclusions in a thin fluid sheet.","core_discovery":"The central discovery is that the solvent's effect can be folded into an anisotropic diffusion tensor: a self term $D_{ii} = D^{\\parallel}_i I + (D^{\\perp}_{ii} - D^{\\parallel}_i) n_i n_i$ and a pair term $D_{ij} = D^{\\perp}_{ij} n_i n_j$ for $i \\neq j$, where $n_i$ is the local membrane normal. The out-of-plane coefficients $D^{\\perp}_{ij}$ are obtained by solving the Stokes equations with Gaussian velocity or stress patches applied on the membrane surface, for three geometries—a single planar membrane, parallel planar membranes, and a spherical vesicle—and integrating the resulting response over each particle's area. The in-plane mobility uses the standard formula for diffusion of a cylindrical inclusion in a fluid sheet. In overdamped Langevin simulations of a $0.5\\,\\mu\\mathrm{m}$ membrane patch, the equilibrium undulation spectrum follows the continuum relation $\\langle h_q h_q^* \\rangle/L^2 = kT/(\\kappa (qL)^4)$, and the two relaxation branches—the slow inter-leaflet slipping mode and the fast hydrodynamic mode—match the continuum dispersion relation given by Eq. (34). For a membrane near a wall, inverting the friction matrix from the parallel-membrane Stokes solution reproduces the continuum bound-membrane dispersion relation of Eq. (36), including an additional slow timescale when the wall is close.","pith_inferences":["Editorial inference: because the pair coupling is restricted to normal-normal components, in-plane hydrodynamic interactions—shear through the membrane and through the solvent—remain outside the model; phase separation or protein aggregation dynamics would need a generalized tensor.","Editorial inference: the Stokes derivation assumes an incompressible, no-slip membrane while the simulated model is compressible; the close match in the tested wavelength range suggests the correction is small there, but softer or highly compressed membranes could show a deviation.","Editorial inference: the length scale $\\alpha$ is a free parameter (the effective particle size for the fluid); since results are stable when $\\alpha$ changes by a factor of ten, one could calibrate $\\alpha$ from a single explicit-solvent simulation and transfer it across resolutions, which the paper does not itself demonstrate.","Editorial inference: the same tensors provide a way to test the paper's scaling argument that out-of-plane solvent dissipation grows as $\\eta R^3/\\kappa$ and eventually dominates budding kinetics; driving a membrane patch with time-dependent curvature forces and comparing relaxation rates would expose that crossover."],"forward_implications":["Solvent-free particle-based membrane models can run with a $0.5\\,\\mathrm{ns}$ timestep while preserving realistic undulation kinetics, because the hydrodynamic coupling is only a pair-list update rather than a solvent or grid solve.","The continuum free-standing dispersion relation is reproduced without fitting the solvent response; the only adjusted parameter is the inter-leaflet friction, which mostly controls the slow slipping mode.","Wall-induced hydrodynamic slowdown emerges from the same tensors: taking the inverse of the friction matrix from the parallel-membrane solution reproduces the bound-membrane dispersion relation, with a clear separation of timescales for wall distances below about $20\\,\\mathrm{nm}$.","The range of hydrodynamic interactions controls non-equilibrium kinetics: longer cut-offs make the largest undulation mode relax faster toward equilibrium, indicating that long-range coupling sets large-scale timescales.","Because the spherical-vesicle solutions converge to the planar results even for small vesicle radii, the planar diffusion tensors remain usable for moderately curved membranes."],"supporting_citations":[{"why":"Supplies the particle-dimer membrane model and force-field parameters used in all simulations.","marker":"[25]"},{"why":"Provides the continuum elastic-membrane dispersion relation with inter-leaflet friction and density fluctuations that the free-standing results are compared against.","marker":"[6]"},{"why":"Provides the continuum dispersion relation for a membrane held near a wall, the reference for the wall simulations.","marker":"[73]"},{"why":"Supplies the overdamped Langevin update and fluctuation-dissipation relations used to generate particle trajectories.","marker":"[78]"},{"why":"Supplies the formula for in-plane mobility of a particle in a thin fluid sheet, used for the parallel component of the self-diffusion tensor.","marker":"[63, 64]"},{"why":"Describes long-range hydrodynamic effects in membranes and the Green's-function solutions that this work extends by imposing membrane incompressibility.","marker":"[44]"},{"why":"Gives the approximate method used to generate correlated random displacements without expensive matrix decompositions.","marker":"[96]"}],"fun_headline_variants":["Stokes tensors add realistic hydrodynamics to solvent-free membranes","Undulation speeds match continuum via Stokes-based friction tensors","Solvent-free membranes get hydrodynamics from Stokes-based tensors","Anisotropic Langevin dynamics restore hydrodynamics to membranes","Stokes solutions add hydrodynamics without extra cost to membrane models"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that each pair of particles feels an independent solvent push along the local membrane normal, spread over a Gaussian patch, while the membrane itself is treated as a no-slip, incompressible sheet; in-plane and many-body hydrodynamic effects are left out.","fun_headline_variants_meta":{"raw":{"variants":["Stokes tensors add realistic hydrodynamics to solvent-free membranes","Undulation speeds match continuum via Stokes-based friction tensors","Solvent-free membranes get hydrodynamics from Stokes-based tensors","Anisotropic Langevin dynamics restore hydrodynamics to membranes","Stokes solutions add hydrodynamics without extra cost to membrane models"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000568,"raw_usage":{"total_tokens":2711,"prompt_tokens":988,"completion_tokens":1723,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":604,"completion_tokens_details":{"reasoning_tokens":1639}},"tokens_in":604,"tokens_out":1723,"duration_ms":12611,"temperature":1.0,"reasoning_tokens":1639,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T04:42:52.741042+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Repeat the same membrane model with explicit solvent or a grid-based hydrodynamic solver and compare the undulation dispersion relations, especially at short wavelengths and small particle separations; a systematic mismatch there would show where the normal-only pairwise tensor misses in-plane or many-body coupling. A second check is to use a much softer membrane, since the derivation's incompressible-boundary assumption should break down more visibly at high q.","supporting_citations":[{"cited_title":"S EIFERT , Phys","cited_arxiv_id":null,"evidence_quote":"Provides the continuum dispersion relation for a membrane held near a wall, the reference for the wall simulations."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the overdamped Langevin update and fluctuation-dissipation relations used to generate particle trajectories."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Describes long-range hydrodynamic effects in membranes and the Green's-function solutions that this work extends by imposing membrane incompressibility."},{"cited_title":"G EYER and U","cited_arxiv_id":null,"evidence_quote":"Gives the approximate method used to generate correlated random displacements without expensive matrix decompositions."}],"review_version":1}