{"id":"18474ff9-dc5a-4ea7-8b93-56693c230667","arxiv_id":"2508.06599","paper_version":6,"verdict":"ACCEPT","confidence":"HIGH","novelty_score":7.0,"correctness_risk":"low","formal_verification":"none","parameter_count":2,"one_line_summary":"At steady state, the concentration of a fully ligand-loaded scaffold is a single-peaked function of total scaffold for any number of independently binding ligands, while some partial complexes can have multiple peaks.","lead":"This paper proves that when several ligand types bind independently to a shared scaffold, the fully loaded complex rises then falls as scaffold is added, with exactly one peak. It also shows that partially loaded complexes can rise and fall more than once, which matters for antibody and gene-circuit design.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"No significant objection identified: Theorem 2 is internally sound; the main caveat (independent binding) is explicit and does not threaten the conditional claim.","rationale":"The reader's ACCEPT is well supported. I verified the main algebraic steps and found no error. The weakest assumption is indeed the independent-binding product form, but the paper declares it and includes Remark 9, so it is a modeling scope limitation rather than a hidden flaw. Hence no change to the verdict.","tokens_in":13535,"tokens_out":12341,"duration_ms":141504,"concrete_test":"As a verification step, independently recompute S(X)=Σ K_iX/(U_i(X)^2+K_iX) and d log F/dX for a parameter set like Figure 8 (K_i=1, A_1=A_2=0.1, A_3=100) by solving the three quadratics (10) on a fine X-grid; check that S crosses 1 exactly once and that F's numerical maximizer matches the sign change. If the crossing count differs, the proof's monotonicity claim for T_i should be rechecked.","verdict_should_be":"UNCHANGED","load_bearing_attack":"I re-derived the proof of Theorem 2. Equations (9)-(14) are correct: (9) follows from (7)-(8); the root U_i>1 is unique and smooth; U_i'<0 in (12); T_i' >0; endpoint limits T_i→0 and T_i→1 are correct; and the intermediate value argument with S going from 0 to m gives exactly one crossing for m≥2. The only substantive limitation—independent binding and the closed-system conservation framing—is stated up front as the model assumption and explicitly discussed in Remark 9. It constrains application to systems such as the CRISPRa example, but it does not create a gap in the theorem. No load-bearing mathematical concern found.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper analyzes a reversible mass-action model in which m ligand types bind independently to a common scaffold, with species Y_I for I ⊆ [m]. The authors prove (Theorem 1) that each positive stoichiometric compatibility class contains a unique positive steady state that is globally asymptotically stable, using detailed balance and the absence of critical siphons. The central result (Theorem 2) shows that, for fixed positive ligand totals and m ≥ 2, the steady-state concentration of the fully bound complex Y_[m] is a differentiable function of total scaffold X = Y_tot with a unique maximizer: it increases strictly up to X* and decreases after. The proof derives a quadratic equation for U_i = 1 + K_i a_i, establishes monotonicity of T_i = K_i X / (U_i^2 + K_i X), and applies the intermediate value theorem to S(X) = Σ T_i. Theorem 3 gives monotonicity of free scaffold (increasing), free ligands (decreasing), and singly bound complexes (increasing). The paper also shows that partially bound complexes can have multiple maxima, with explicit numerical examples for m = 3 and m = 4. The main theorem is conditional on the independent-binding assumption and the closed-system conservation framing, both clearly stated in Remark 9.","tokens_in":13757,"tokens_out":6818,"duration_ms":77716,"significance":"If accepted, this paper provides a rigorous, parameter-free explanation of biphasic dose-response in scaffold systems, directly relevant to bispecific/trispecific antibody design and CRISPRa synthetic biology. The proof is unusually self-contained: the key derivative computation (Eq. (14)) is transparent, and the uniqueness argument via a strictly increasing S(X) is elegant. The paper also contributes a useful cautionary result: intermediate partial complexes need not be uniquely biphasic, with reproducible numerical examples. The explicit correction to a published formula (Remark 8) is a small but concrete contribution. The assumptions are clearly delimited, no data fitting is involved, and the central theorem is a falsifiable prediction of the model.","major_comments":[],"minor_comments":[{"comment":"The limitations of the model are explicitly acknowledged: independent binding may fail (e.g., facilitated promoter binding) and the closed-system framework omits production, degradation, and dilution. This is a scope limitation, not an internal inconsistency; the conditional claim in Theorem 2 is unaffected. The authors should keep this remark, as it correctly prevents overgeneralization.","section":"Remark 9"},{"comment":"The phrase 'each stoichiometric compatibility class contains a unique steady state' is imprecise. Theorem 1 concerns 'each positive stoichiometric compatibility class' and its positive steady state. The abstract and introduction should be aligned with the formal statement to avoid confusion about boundary behavior.","section":"Abstract and Section 1"},{"comment":"Typo: 'intutively' should be 'intuitively'.","section":"Remark 6"},{"comment":"The caption states that the vertical axis is scaled to the largest local maximum but does not give the scale factor. Adding the factor would aid reproducibility of the numerical example.","section":"Figure 10 caption"},{"comment":"In the endpoint-limit argument as X → 0+, the boundedness of U_i is established via U_i - 1 < α_i. This step is correct but slightly compressed; a parenthetical reminder that (U_i + K_i X)/U_i > 1 would improve readability.","section":"Section 5, proof of Theorem 2"}],"recommendation":"minor_revision","confidential_remarks":"The paper is a good fit for the journal. The stress-test concern about the independent-binding assumption is explicitly addressed in Remark 9 and does not undermine the conditional theorem. The central derivation is sound; the remaining issues are presentation/local. I recommend minor revision to fix the typos and to make the abstract's compatibility-class wording precise."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"You should know this paper by Sontag (arXiv:2508.06599) is a genuinely useful piece of mathematical pharmacology. It proves what practitioners have assumed for m=2 and guessed for m>2: for any number of independently binding ligands, the steady-state concentration of the fully bound scaffold complex is strictly biphasic in total scaffold, with exactly one peak. The proof is elementary and I checked the key steps—equations (9)–(14), the construction of T_i, the endpoint limits—and they are solid. The theorem is parameter-free: it holds for all positive K_i and A_i,tot, so it is a structural result, not a fitting artifact.\n\nWhat's new: the general m≥2 theorem, the explicit multiphasic behavior of partial complexes (two or three local maxima), and the sharp observation that biphasic shape is not a trivial consequence of multiplying an increasing and a decreasing function (Remark 7). That remark alone is worth teaching.\n\nThe soft spots are real but modest. The independent-binding assumption is the main one; the author flags it in Remark 9 and gives the CRISPRa example where it may fail. If binding is cooperative or sequential, Theorem 2 does not apply. The closed-system conservation framing also ignores production and dilution. But these are stated assumptions, not hidden gaps. The numerical examples are not accompanied by code, but the formulas specify them completely; reproducibility is a minor issue. The claimed correction of a published m=2 formula in [5] should be double-checked by any user, but it looks right.\n\nThere are no data, which is appropriate for a theory paper. The citation pattern is honest; self-citations are to standard CRN theory and the author's prior work, and the m=2 partial prior results are acknowledged.\n\nWho should read this: pharmacologists modeling bispecific or trispecific antibodies, synthetic biologists working with scaffold RNAs, and theorists wanting a clean example of dose-response shape analysis. It deserves a serious referee; I would send it to review and expect it to be accepted after minor revisions. If you are in this area, cite it—it is the missing rigorous backbone for scaffold dosing.","headline":"A rigorous, parameter-free proof that the fully bound scaffold complex is biphasic in total scaffold for any number of independently binding ligands, plus a surprising multiphasic caveat for partial complexes.","tokens_in":14248,"tokens_out":1772,"would_cite":true,"duration_ms":19926,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["92C42","34D23","37N25"],"pacs":[],"model":"deepseek-v4-flash","headline":"This paper proves that under independent binding, the fully bound scaffold complex rises to a unique maximum and then falls as total scaffold increases, for any number of ligand types greater than one.","keywords":["scaffold proteins","independent binding","biphasic dose response","bispecific antibodies","trispecific antibodies","chemical reaction networks","detailed balance","global stability"],"falsifier":"Search the parameter space of the $m=2$ mass-action model, or run a well-mixed in vitro binding assay with two ligands and a scaffold satisfying independent binding, for a case where the fully bound complex has two distinct local maxima as total scaffold varies; Theorem 2 says none exists, so a single such numerical or experimental curve would refute the central claim.","tokens_in":13400,"feed_emoji":"💊","tokens_out":7446,"duration_ms":82753,"temperature":0.7,"pith_summary":"The paper asks a quantitative question about scaffolds: as you add more scaffold, how does the amount of fully loaded complex behave? The answer, proved for any number $m\\ge 2$ of independently binding ligand types under mass-action kinetics, is that it always rises to a single maximum and then falls. This biphasic dose response is the mathematical reason bispecific antibody drugs have an optimal concentration, and the proof shows the same holds for trispecific antibodies and CRISPRa scaffolds. The paper also proves that every stoichiometric compatibility class has a unique, globally stable steady state, and that partially loaded complexes can behave more intricately, with multiple peaks.","feed_headline":"Multi-ligand scaffolds peak at exactly one dose","feed_subtitle":"A single optimal scaffold dose exists for bispecific and trispecific antibody designs, not just in special cases.","key_machinery":"The load-bearing machinery is the detailed-balance product form $y_I = y_\\emptyset \\prod_{j\\in I} K_j a_j$, which turns the conservation laws into $m$ decoupled quadratic equations for $U_i=1+K_i a_i$. The scalar function $S(X)=\\sum_{i=1}^m T_i(X)$ with $T_i(X)=K_iX/(U_i(X)^2+K_iX)$ then carries the entire argument: each $T_i$ is strictly increasing in $X$, the sum runs from $0$ to $m$, and the unique solution of $S(X)=1$ gives the unique maximizer of the fully bound complex.","core_discovery":"The central discovery is Theorem 2. Fix $m\\ge 2$ ligands with positive totals $A_{1,\\mathrm{tot}},\\dots,A_{m,\\mathrm{tot}}$; under independent binding, for each total scaffold concentration $X=Y_{\\mathrm{tot}}$ there is a unique positive steady state. The concentration $F(X)$ of the fully bound complex $Y_{[m]}$ is differentiable and has exactly one critical point $X^*$, with $F'(X)>0$ before and $F'(X)<0$ after. The proof reduces the steady-state equations to decoupled quadratics for $U_i=1+K_i a_i$ and expresses $F$ as $X\\prod_j (U_j-1)/U_j = (\\prod_j \\alpha_j)X\\big/\\prod_j (U_j+K_jX)$. Its logarithmic derivative is $(1-S(X))/X$, where $S(X)=\\sum_i K_i X/(U_i^2+K_iX)$; each summand is stri","pith_inferences":["Testable extension: if binding is made cooperative or allosteric, the product form breaks, and the predicted signature is loss of the single peak; a scaffold titration with a ligand that changes another ligand's affinity could distinguish the independence regime.","The paper's mechanism for partial complexes suggests a design heuristic for synthetic scaffolds: partial complexes can show multiple peaks whenever omitted ligands are abundant on separated scales, so using the fully bound complex as the readout avoids this ambiguity.","Because the model is closed, applying the theorem to live cells requires the additional assumption that production, degradation, and dilution are slow relative to binding; the paper leaves open how those processes reshape the curve.","A natural mathematical extension is time-varying scaffold input: the monotonic crossing function $S(X)$ may also control transient ordering of dose responses, but the paper does not address dynamics."],"forward_implications":["For bispecific antibody drugs ($m=2$), this is a rigorous proof that a unique optimal antibody concentration exists, justifying dosing strategies aimed at maximizing the fully bound ternary complex.","For any $m\\ge 2$, adding scaffold beyond the optimum is guaranteed to reduce the fully bound complex, so overdosing trispecific antibodies or CRISPRa scaffolds is mathematically certain to be counterproductive.","Intermediate partially bound complexes can have two or more local maxima, so dose-response experiments should measure the fully bound complex rather than a partial complex when looking for the unique optimum.","Free scaffold and singly bound complexes increase monotonically with total scaffold, while free ligand concentrations decrease; singly bound complexes act like capture curves approaching $A_{i,\\mathrm{tot}}$.","For a single ligand ($m=1$) there is no finite optimal scaffold dose; the fully bound complex increases monotonically to a limiting value."],"supporting_citations":[{"why":"Supplies the standard theory of detailed-balanced mass-action systems: a unique positive equilibrium in each stoichiometric class and local asymptotic stability.","marker":"[7]"},{"why":"Gives the persistence criterion via absence of critical siphons, used to rule out boundary omega-limit points in the global convergence proof.","marker":"[4]"},{"why":"Provides the result that complex-balanced trajectories either converge to their equilibrium or accumulate on the boundary, combined with persistence.","marker":"[14]"},{"why":"Also cited for the complex-balanced convergence result used in the proof of global stability.","marker":"[3]"},{"why":"Provides the explicit $m=2$ three-body binding formula that the paper recovers, anchoring the biphasic response in existing drug-design models.","marker":"[5]"},{"why":"Prior work on bispecific antibody dynamics and identifiability that motivates the $m=2$ case and its optimal-dose question.","marker":"[11]"}],"fun_headline_variants":["Unique scaffold dose maximizes full binding","One optimal scaffold level, proven","Scaffold binding peaks at a single dose","Multi-ligand scaffold: one peak response","Proof of unimodal scaffold dose response"],"cache_read_input_tokens":2816,"weakest_assumption_plain":"The result rests on independent binding—each ligand's binding and unbinding rates are the same no matter what else is already on the scaffold—and on a closed system with no production, degradation, or dilution; if binding is context-dependent or the system is open, the unique-maximum conclusion is not guaranteed.","fun_headline_variants_meta":{"raw":{"variants":["Unique scaffold dose maximizes full binding","One optimal scaffold level, proven","Scaffold binding peaks at a single dose","Multi-ligand scaffold: one peak response","Proof of unimodal scaffold dose response"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":7.7e-05,"raw_usage":{"total_tokens":958,"prompt_tokens":739,"completion_tokens":219,"prompt_tokens_details":{"cached_tokens":640},"prompt_cache_hit_tokens":640,"prompt_cache_miss_tokens":99,"completion_tokens_details":{"reasoning_tokens":156}},"tokens_in":99,"tokens_out":219,"duration_ms":189477,"temperature":1.0,"reasoning_tokens":156,"cache_read_input_tokens":640,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-05T22:43:28.463360+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Search the parameter space of the $m=2$ mass-action model, or run a well-mixed in vitro binding assay with two ligands and a scaffold satisfying independent binding, for a case where the fully bound complex has two distinct local maxima as total scaffold varies; Theorem 2 says none exists, so a single such numerical or experimental curve would refute the central claim.","supporting_citations":[{"cited_title":"Feinberg.Foundations of Chemical Reaction Network Theory, volume 202 ofApplied Mathemati- cal Sciences","cited_arxiv_id":null,"evidence_quote":"Supplies the standard theory of detailed-balanced mass-action systems: a unique positive equilibrium in each stoichiometric class and local asymptotic stability."},{"cited_title":"Angeli, P","cited_arxiv_id":null,"evidence_quote":"Gives the persistence criterion via absence of critical siphons, used to rule out boundary omega-limit points in the global convergence proof."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the result that complex-balanced trajectories either converge to their equilibrium or accumulate on the boundary, combined with persistence."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Also cited for the complex-balanced convergence result used in the proof of global stability."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the explicit $m=2$ three-body binding formula that the paper recovers, anchoring the biphasic response in existing drug-design models."},{"cited_title":"Sadeghi, I","cited_arxiv_id":null,"evidence_quote":"Prior work on bispecific antibody dynamics and identifiability that motivates the $m=2$ case and its optimal-dose question."}],"review_version":1}