{"id":"d2eb8572-1eb7-492c-b769-5696b5cbc723","arxiv_id":"1908.08309","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":5,"one_line_summary":"Coarse-grained simulations of tethered bivalent nanobody constructs show that nanobody shape, linker length, and weak wall attraction control the kinetics of the second binding event, with a survival-time protocol that extracts on/off rates.","lead":"This study builds two simplified computer models of a nanobody-polymer-nanobody drug and simulates how the free arm moves near a surface. It shows that nanobody shape and weak attractions with the surface change how often and how long the free arm is close enough to bind a second target.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Absolute rate constants rest on an unspecified LJ-to-physical-time conversion: with only sigma and epsilon set, and no mass unit or Langevin damping coefficient reported, the ns/µs values in Section II.C and Tables II-VI are not reproducible.","rationale":"The reader's weakest assumption identifies the missing mass unit and time calibration for the kinetic constants. My independent reading reaches the same conclusion, sharpened by the observation that for overdamped Langevin dynamics even the mass unit is insufficient—the damping coefficient is an additional required parameter, and no value is reported anywhere in the manuscript or its appendix. This is the most load-bearing concern because it affects every numerical kinetic result in the paper, not merely one figure or table. The relative comparisons that support the qualitative claims about shape and weak attractions are indeed invariant under a global rescaling, so the internal physics may remain valid; however, the stated purpose of the paper is to establish a computational protocol for drug design, and a protocol that cannot convert its simulation time to physical time cannot produce the quantitative kinetic constants it promises. The concern is addressable by reporting the missing parameters or by providing an independent diffusion-coefficient calibration, which is why I do not recommend rejection. The reader's CONDITIONAL verdict remains appropriate, so I recommend UNCHANGED.","tokens_in":31611,"tokens_out":11185,"duration_ms":114151,"concrete_test":"Re-run one SPH 30-mer flight/residence simulation with identical σ, ε, reduced temperature, and all interaction parameters, but with the Langevin damping coefficient (Tdamp) changed by a factor of 10, adjusting the timestep proportionally to maintain numerical accuracy. If the resulting mean flight and residence times in reported ns change by the same factor, the absolute rates are controlled by an unstated parameter and the conversion is not reproducible; if they remain unchanged, the authors must still supply the mass unit and timestep used to define the reduced time unit.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central quantitative output of Section II.C—the flight and residence times in ns, and hence the on/off rates in Eqs. (9)–(10) and Figures 8–11 and Tables II–VI—depends on converting Lennard-Jones simulation time to physical time. The Methods (Section I.A) specify the length unit σ = 3.5 Å and the energy unit ε = 100 K, and say that bead masses are set to 1, but they give no mass unit, no timestep, and no Langevin damping coefficient. This is not a cosmetic omission. In an overdamped Langevin simulation performed with LAMMPS in reduced units, the physical time unit is not fixed by σ and ε alone; the reduced friction coefficient (LAMMPS Tdamp) sets the diffusion coefficient and therefore the absolute timescale. Even taking the inertial timescale, the time unit would be σ√(m/ε), which remains unknown until the physical mass of the 'mass = 1' beads is stated. The paper's claim to quantify encounter/dissociation kinetics in physical units is thus underdetermined. Relative trends (SPH versus SBCG, repulsive versus attractive wall) are unaffected by a global time rescaling, but the protocol's purpose is to provide design-relevant kinetic constants, and those constants are exactly what cannot be reproduced without the missing calibration.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper introduces two coarse-grained models of a polymer-linked bivalent nanobody construct—a spherical model (SPH) and a shape-based coarse-grained model (SBCG) derived from an atomistic simulation—and uses Langevin dynamics to study the thermodynamics and kinetics of the free nanobody relative to a tethering wall. Umbrella sampling is used to compute potentials of mean force along two reaction coordinates, and long trajectories are analyzed by classifying paratope z-coordinate excursions into flight and residence events relative to a threshold. Survival probabilities of these events are then used to extract encounter and dissociation rates via Eqs. (9)-(10), with a focus on how linker length, nanobody shape, and weak nonspecific wall attractions affect these rates. The central claim is that this mesoscale protocol can quantify the kinetic determinants of the second, cooperative binding step and that the SBCG model captures shape-dependent features missed by the simpler SPH model.","tokens_in":31925,"tokens_out":6413,"duration_ms":61102,"significance":"If the quantitative timescale issue is resolved, this is a useful contribution to mesoscale computational design of multivalent therapeutics. The PMF calculations are extensive and the distinction between short-time diffusive t^{-1/2} survival and exponential rare-event tails is clearly supported by the data. The paper's strengths include the direct measurement of survival statistics from long trajectories, the transparent first-passage formulas, and the explicit comparison of two coarse-graining schemes. The main limitation is that the reported absolute kinetic constants in nanoseconds/microseconds are not reproducible without specifying the mapping from Lennard-Jones reduced time to physical time; the relative trends among models and wall types are robust to a global time rescaling, but the design-oriented absolute rates are not.","major_comments":[{"comment":"The paper reports all flight/residence times and rates in ns/µs, but the conversion from Lennard-Jones reduced time to physical time is never specified. Section I.A states σ = 3.5 Å, ε = 100 K, and bead masses of 1, but gives no mass unit, no timestep, and no Langevin damping/friction coefficient. In an overdamped Langevin simulation, the reduced friction coefficient (LAMMPS Tdamp) sets the diffusion coefficient and therefore the physical timescale; for an inertial mapping one would need the physical bead mass. Without this information, the numerical values in Tables II-VI and Figures 8-11, as well as the 0.5 ns minimum event duration in Section II.C, cannot be reproduced or compared with experiments. The authors should provide the full reduced-to-physical time conversion protocol, or explicitly state that all times are in reduced units and remove the physical units throughout.","section":"I.A, II.C, Eqs. (9)-(10), Tables II-VI"},{"comment":"The SBCG flight/residence time data are obtained from a single 15 µs simulation per linker length and are reported without error bars (Table VI). The central conclusion that the SBCG model yields systematically shorter times than the SPH model (e.g., ~50 ns versus ~80 ns flight time for the 30-mer) rests on this comparison. Without a statistical uncertainty estimate, it is difficult to assess whether the observed differences are significant. The authors should provide error estimates via block averaging, bootstrap, or replicate simulations for the SBCG systems, as is already done for the SPH data.","section":"I.E, Table VI, Section II.C"}],"minor_comments":[{"comment":"The symbol z0 is used both for the tethering height introduced in Section II.B (z0 = 7.5σ) and as the lower integration limit/threshold in Eqs. (13)-(16). Since the flight/residence threshold is zth = 3.5σ, using z0 in these equations is confusing and should be replaced with zth or a distinct symbol.","section":"II.B-II.C, Eqs. (13)-(16)"},{"comment":"The statement that the energy unit is ε = 100 K is imprecise; it should read ε/k_B = 100 K or the numerical value in joules should be given. Also, no timestep is reported anywhere, even though the manuscript states simulation lengths in time steps and in physical time units.","section":"I.A"},{"comment":"The caption appears to contain duplicated panel labels: the text references panels (A), (B), (C), (D), but the displayed caption lines show '(A) (B)' twice. Please verify the panel labeling in the figure and caption.","section":"Figure 5 caption"},{"comment":"The phrase 'free/constrained dynamics (100 ns − µs)' should be written as '100 ns to µs' or '100 ns - 1 µs' to avoid ambiguity.","section":"Abstract"}],"recommendation":"major_revision","confidential_remarks":"The time-calibration gap is the main blocker for the paper's quantitative claims; it is fixable and does not undermine the relative trends or the methodological framework. I also note the absence of a statement about code/data availability, which would be helpful for a protocol paper. The lack of error bars for the SBCG data should be addressed in revision, as it bears on one of the central comparative conclusions."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Two things to know. First, this is a usable mesoscale protocol paper: the SPH vs SBCG comparison is genuinely useful, and the finding that molecular shape matters for near-wall kinetics is worth having. Second, the headline numbers—the absolute on/off rates in ns and µs—are not reproducible as reported, because the paper never specifies how Lennard-Jones time is mapped to physical time.\n\nWhat's actually new: the systematic comparison between a spherical-bead nanobody model and a shape-based coarse-grained model for tethered diabodies, and the observation that the SBCG shape flattens the xy PMF and shortens flight times for short linkers. The flight/residence decomposition itself is standard survival analysis, but turning it into a protocol for estimating second-binding kinetics is a reasonable methodological contribution. The umbrella sampling PMFs look well converged, and the authors show a convergence check for flight times, which is more than many papers do.\n\nThe main soft spot is the one above: Methods give σ=3.5 Å, ε=100 K, masses set to 1, but no mass unit, no Langevin damping coefficient, no timestep. In overdamped Langevin dynamics, the physical time scale is set by the friction/damping, not just by σ and ε. So the ns and µs values in the tables and figures are under-determined. This is not a cosmetic omission. The relative trends—shape effects, wall attraction effects, linker-length dependence—are unaffected by a global time rescaling, so the qualitative conclusions likely survive. But the protocol's purpose is to provide design-relevant kinetic constants, and those constants cannot be reproduced from the paper as written. The authors need to either supply the missing parameters or explicitly reframe the results as relative.\n\nWeaker issues, in order: the SBCG kinetics come from a single 15 µs run per linker length with no error bars, whereas the SPH data have three runs and error bars; no code or data are released; and the SBCG wall parameters differ from the SPH ones in ways that make the model comparison not perfectly controlled, though the paper does note this.\n\nOverall, it's an honest, competent piece of work. It deserves a serious referee, and I'd send it out with a request to address the time calibration and error bars. If the authors can fix those, it becomes a solid methods paper for the biophysics coarse-graining community.","headline":"A credible mesoscale protocol for studying second-binding kinetics, but the absolute rate constants in ns/µs are not reproducible because the simulation-time-to-physical-time conversion is never specified.","tokens_in":32431,"tokens_out":3065,"would_cite":true,"duration_ms":31414,"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 kinetics of the second binding step in bivalent nanobody constructs can be extracted from coarse-grained simulations by splitting paratope trajectories into flight and residence events.","keywords":["bivalent nanobody","avidity","coarse-grained simulation","flight-residence kinetics","potential of mean force","second binding rate","nanobody shape","Langevin dynamics"],"falsifier":"An explicit-solvent atomistic simulation of a single nanobody tethered near a wall, run long enough to accumulate residence events, would provide an independent mean residence time; if it disagrees with the corresponding SBCG value of a few nanoseconds by more than the statistical error, the coarse-grained time conversion is wrong.","tokens_in":31405,"feed_emoji":"🧬","tokens_out":8394,"duration_ms":82263,"temperature":0.7,"pith_summary":"The paper's aim is to make bivalent nanobody constructs designable at the mesoscale: given a polymer-linked pair of nanobodies, it wants to predict from coarse-grained simulation how readily the free nanobody encounters and then leaves the target surface, the kinetic step that controls avidity. The authors introduce two coarse-grained models, one with rigid spherical nanobodies and one with shape-based flexible nanobodies, and split each simulated paratope trajectory into \"flight\" stretches above a threshold height and \"residence\" stretches below it. From the survival statistics of those stretches they extract an on-rate and an off-rate for the wall encounter. They report that nanobody shape and weak nonspecific attractions of 1–2 $k_B T$ measurably change these rates, and that rare long events, not averages, dominate the tail statistics. If this protocol holds up, linker lengths and nanobody geometries can be screened computationally before synthesis.","feed_headline":"Coarse-grained runs predict how fast nanobodies find a second target","feed_subtitle":"Simulated flight and residence times yield the on/off rates of the second binding step.","key_machinery":"The machine that carries the argument is the thresholded paratope-height trajectory. The authors define flight and residence domains by $z \\ge z_{\\mathrm{th}}$ and $z < z_{\\mathrm{th}}$, require events to last at least 0.5 ns, and convert the event-time histograms into survival probabilities. The identity connecting the two is first-passage: the survival probability in a domain is the integral of the exit-time distribution, so the inverse mean exit time is the escape rate (Eqs. 7–10). A second piece of machinery is the set of potentials of mean force from umbrella sampling along the vertical and lateral reaction coordinates; these show the entropic barrier a shaped nanobody feels near the wall, explaining why the simple spherical model misrepresents wall encounters.","core_discovery":"The central claim is that the encounter and dissociation kinetics of the free nanobody with the target surface, the determinants of avidity in a bivalent construct, can be read directly from a coarse-grained trajectory. The operational definition is a height threshold $z_{\\mathrm{th}}=3.5\\sigma$ for the paratope above the tethering wall; consecutive frames below that threshold are residence events, above it flight events, with events shorter than 0.5 ns discarded to avoid recrossing noise. The survival probabilities $S_f(t)$ and $S_r(t)$ of the two event types are related to exit times by $P_f=-dS_f/dt$ and $P_r=-dS_r/dt$, so the inverse mean exit times give $k_{\\mathrm{on}}=[\\int_0^\\infty S_f(t)\\,dt]^{-1}$ and $k_{\\mathrm{off}}=[\\int_0^\\infty S_r(t)\\,dt]^{-1}$. Applied to rigid-sphere and shape-based models, the protocol yields $\\langle t_f\\rangle$ of order 100 ns and $\\langle t_r\\rangle$ of order 2–3 ns, with the shape-based model showing shorter times and higher on-rates, and weak wall attraction raising $k_{\\mathrm{on}}$ and lowering $k_{\\mathrm{off}}$.","pith_inferences":["A direct extension would be to feed the extracted $k_{\\mathrm{on}}$ and $k_{\\mathrm{off}}$ into a two-site kinetic model of avidity and convert the per-encounter rates into a predicted fold-gain in effective affinity; the paper does not carry out that conversion.","The threshold $z_{\\mathrm{th}}=3.5\\sigma$ and the 0.5 ns cutoff are free choices; one could test whether the ranking of designs is invariant under modest changes in these choices, turning the method into a robust first-passage observable.","I infer the method should transfer to the immune-synapse geometry by replacing the single wall with two parallel walls and counting residence events in the inter-membrane gap, where experimental rebinding measurements would provide a direct check.","Because the shape-based model is built from one crystal structure, the protocol could be run on multiple nanobody sequences or epitopes to see whether shape alone, not chemistry, sets the ranking."],"forward_implications":["Linker length can be screened in silico: the protocol predicts a non-monotonic flight time with a minimum near a linker length comparable to the nanobody size, about 20 monomers.","Weak nonspecific surface attraction of 1–2 $k_B T$ becomes a design variable: it raises the second-binding on-rate and lengthens residence times.","Shape-based coarse graining is needed for wall-proximal kinetics; the rigid-sphere model overestimates flight and residence times for short linkers.","Rare, long events are quantitatively significant: exponential tails set tens-of-nanosecond residence scales and microsecond flight scales, which matters in low-copy geometries such as the immune synapse."],"supporting_citations":[{"why":"Supplies the shape-based coarse-graining method used to build the flexible SBCG nanobody model.","marker":"[33]"},{"why":"Provides the PEG persistence length used to set the bending stiffness of the linker.","marker":"[41]"},{"why":"Supplies the Langevin dynamics engine in which all coarse-grained simulations are run.","marker":"[42]"},{"why":"Gives the Gaussian tethered-polymer wall distribution used as the baseline model for the paratope z-distribution.","marker":"[47]"},{"why":"Provides the first-passage formalism that connects survival probabilities to exit-time distributions and rates.","marker":"[50]"},{"why":"Frames the sequential-binding picture of bivalent molecules that motivates studying the second-binding encounter step.","marker":"[51]"},{"why":"Supplies the crystal structure used as the starting point for the atomistic trajectory that feeds the SBCG construction.","marker":"[34]"}],"fun_headline_variants":["Coarse-grained simulations compute nanobody on/off rates","Nanobody dwell and flight times set avidity rates","Mesoscale protocol predicts nanobody–surface binding kinetics","Nanobody flexibility drives binding kinetics in simulations"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The reported nanosecond rates rest on an implicit conversion of Lennard-Jones simulation time to real time: the paper sets $\\sigma=3.5$ Å and $\\epsilon=100$ K, leaves bead masses and time units uncalibrated, and therefore the absolute values of $k_{\\mathrm{on}}$ and $k_{\\mathrm{off}}$ depend on an unstated time mapping.","fun_headline_variants_meta":{"raw":{"variants":["Coarse-grained simulations compute nanobody on/off rates","Nanobody dwell and flight times set avidity rates","Mesoscale protocol predicts nanobody–surface binding kinetics","Nanobody flexibility drives binding kinetics in simulations"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00191,"raw_usage":{"total_tokens":7551,"prompt_tokens":1083,"completion_tokens":6468,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":699,"completion_tokens_details":{"reasoning_tokens":6406}},"tokens_in":699,"tokens_out":6468,"duration_ms":42837,"temperature":1.0,"reasoning_tokens":6406,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T11:42:53.483097+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"An explicit-solvent atomistic simulation of a single nanobody tethered near a wall, run long enough to accumulate residence events, would provide an independent mean residence time; if it disagrees with the corresponding SBCG value of a few nanoseconds by more than the statistical error, the coarse-grained time conversion is wrong.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the shape-based coarse-graining method used to build the flexible SBCG nanobody model."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the PEG persistence length used to set the bending stiffness of the linker."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the Langevin dynamics engine in which all coarse-grained simulations are run."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Gives the Gaussian tethered-polymer wall distribution used as the baseline model for the paratope z-distribution."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the first-passage formalism that connects survival probabilities to exit-time distributions and rates."},{"cited_title":"Proceedings of the National Academy of Sciences, 103(16), 61666171 (2006)","cited_arxiv_id":null,"evidence_quote":"Frames the sequential-binding picture of bivalent molecules that motivates studying the second-binding encounter step."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the crystal structure used as the starting point for the atomistic trajectory that feeds the SBCG construction."}],"review_version":1}