{"id":"877157b2-8840-4fc8-a7df-cd247fed05bf","arxiv_id":"2607.17072","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"Stepwise periodic driving of a Michaelis–Menten enzyme yields a long-time product current that scales as 1/Ω at high frequency and Ω at low frequency, with sign set by the cyclic switch order.","lead":"A theory paper works out exact asymptotic formulas for the net product current produced when an enzyme's reaction rates are switched periodically and abruptly. The results show the switch order sets the current's direction and give simple 1/Ω (fast) and Ω (slow) scaling laws, offering a handle for controlling biochemical reactions with light.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"No significant objection identified: central coefficients verified; van Vleck rigor gap is a caveat, not a demonstrated failure.","rationale":"The paper's central claim is that the long-time current for the three-step protocol obeys Eq. (75) at high Ω and Eq. (83) at low Ω, with sign reversal under reordering. I checked both analytically. For the high-frequency side, inserting the van Vleck generator (66) into Eq. (33) gives the W0 (dynamical) contribution zero because ⟨S|J0|ψ0⟩=0, while the correction matrix produces ⟨S|(−i∂χM)|ψ0⟩=1/126, hence (75). For the low-frequency side, I computed the geometric coefficient from first-order perturbation theory of the 2×2 dominant eigenmodes. With the gauge ∑u=1 and ⟨l|u⟩=1, the overlap derivative for each adjacent pair is ⟨l'_{a+1}|u_a⟩ = i(c_{a+1}+d_{a+1})/(Σ_{a+1}^2 Σ_a)(T_{a+1}S_a − S_{a+1}T_a). Summing cyclically over the three steps in (51)–(53) gives exactly 1/100, so (83) is correct. The only genuine weakness is the lack of a rigorous error bound for the non-Hermitian van Vleck expansion. This is a caveat, not a demonstrated flaw: the protocol is finite-dimensional and bounded, so U(T) is analytic in T and log U(T)/T has a convergent Taylor expansion near T=0; the van Vleck formula is the first-order term of that expansion. The paper's validity discussion (§IV.C.3) gives reasonable criteria and numerics support the asymptotes. The missing eigenvector derivation and the absence of code/data are reproducibility issues; they do not shift the central claim. Hence I do not see a load-bearing objection that would change the conditional verdict.","tokens_in":23629,"tokens_out":24653,"duration_ms":209756,"concrete_test":"Independently expand the exact one-period propagator U(T)=e^{W2 T/3}e^{W1 T/3}e^{W0 T/3} for the protocol in Eqs. (51)–(53) to first order in T around T=0, forming log U(T) and taking −i∂χ log U(T) at χ=0. Verify that the coefficient of T equals 1/126. If it matches, the van Vleck transfer is confirmed for this protocol; if not, Eq. (75) is wrong.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's central high- and low-frequency claims survive scrutiny. Eq. (75) follows correctly from the van Vleck first-order generator (66): the W0 contribution vanishes in |ψ0⟩, and the correction M=(2π/Ω)[[0,0],[1/54(1-e^{-iχ}),0]] yields ⟨S|(-i∂χM)|ψ0⟩=1/126, giving J∞=(1/126)(2π/Ω). I also independently re-derived the low-frequency geometric coefficient via first-order perturbation theory of the 2×2 dominant eigenmodes; the cyclic sum of overlap derivatives gives exactly 1/100, matching Eq. (83). The reader's weakest assumption—that the van Vleck expansion is imported without an error bound for non-Hermitian, piecewise-constant generators—is a legitimate rigor gap, but it is not load-bearing. For finite-dimensional bounded W(t), the one-period propagator U(T) is an analytic function of T in a neighborhood of T=0, so (1/T)log U(T) has a convergent Taylor expansion; the van Vleck formula is the first-order term of that expansion, not merely a formal asymptotic series. The paper's validity criteria in §IV.C.3 (Ω/π≫1; |Re(λ1−λ0)|τ≳1) are consistent with the numerical range shown. The omitted low-frequency eigenvector derivation and the absence of code/data are presentation/reproducibility weaknesses, but no detected error in the central argument.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper develops a Floquet formalism for periodically driven continuous-time Markov processes in the presence of a counting field, with the goal of computing long-time currents and counting statistics. The central formal object is Eq. (25), which expresses the net current at finite time through the effective Floquet generator and the kicked state. The paper then derives asymptotic descriptions in two limits: a high-frequency van Vleck expansion of the effective generator (Sec. III.B, Eqs. (33), (62)–(64)) and a low-frequency geometric-phase formulation (Sec. III.C, Eqs. (44)–(46)). These are applied to a two-state Michaelis–Menten enzyme with a three-step discrete driving protocol (Eqs. (51)–(53)). The main quantitative claims are J∞(Ω) = (1/126)(2π/Ω) + O(Ω^{-2}) at high frequency (Eq. (75)) and J∞(Ω) = (1/100)(Ω/2π) at low frequency (Eq. (83)), with sign reversal when the cyclic order of generators is reversed (Eqs. (76), (84)); both are compared with numerical integration in Fig. 5. An effective-reaction-rate interpretation of the high-frequency regime is given in Eqs. (68)–(70).","tokens_in":23981,"tokens_out":5757,"duration_ms":56067,"significance":"If the claims hold, the paper provides a non-perturbative extension of earlier perturbative treatments of stochastic pumping and geometric currents in chemical reaction networks. The high-frequency effective-generator picture is intuitively appealing and directly usable, while the low-frequency geometric decomposition extends known results to discrete, abruptly switched protocols. The paper has no fitted parameters: the asymptotic coefficients are determined analytically and compared with numerical integration. The explicit criterion for the low-frequency regime (Sec. IV.C.3) and the numerical convergence study in App. A are useful additions. The work is likely to be of interest to researchers in stochastic thermodynamics, Floquet engineering, and biophysical modeling of optogenetically controlled enzymes.","major_comments":[{"comment":"The low-frequency asymptotic coefficient J_dyn = 0 and J_geom = (1/100)(Ω/2π) are stated as 'explicitly calculated', but no calculation is shown. Eqs. (78)–(80) require the dominant left and right eigenvectors of the three 2×2 generators and their χ-derivatives. This is a finite but nontrivial computation, and it is load-bearing for the low-frequency central claim and for the numerical comparison in Fig. 5. I independently checked the coefficient, but the manuscript should make the derivation traceable—either in the main text or an appendix—by giving the eigenvectors and the χ-derivative overlaps that produce the factor 1/100 and the vanishing dynamical contribution.","section":"§IV.A.2 (Eq. (83)) and App. D"},{"comment":"The van Vleck high-frequency expansion, Eqs. (27), (62)–(64), is imported from Hermitian quantum Floquet theory without an error bound or convergence argument for the non-Hermitian, piecewise-constant Markov generator W(χ;t). The validity criterion in §IV.C.3 is a heuristic magnitude estimate (Ω/π ≫ 1) rather than a proof. For finite-dimensional bounded generators, the one-period propagator is analytic near T = 0, so (1/T) log U(T) has a convergent expansion and the gap is likely fillable. I ask the authors to add a short justification (or explicitly state that the expansion is formal and used in the asymptotic sense) so that the central 1/Ω formula is not left as an unproved import.","section":"App. C / §III.B"}],"minor_comments":[{"comment":"Please state the normalization of |ψ0⟩ used in the high-frequency calculation. Eq. (E15) uses the normalization ⟨S|ψ0⟩ = 1; this is not explicit when the vector (3/7, 4/7)^T is introduced, and it is needed to reproduce Eq. (75) from Eq. (33).","section":"§IV.A.1, Eq. (73)"},{"comment":"The numerical simulations are described only qualitatively. Please add a brief numerical-methods paragraph (ODE solver, time-step control, tolerances, number of periods) and, ideally, a data/code availability statement or repository, since the paper claims numerical validation of the two asymptotics.","section":"Fig. 5 / numerical methods"},{"comment":"The Fourier-mode notation W_m, with negative indices, is used heavily. A one-sentence definition of the Fourier decomposition and the index convention (as in Eq. (D6)) before Eq. (64) would improve readability and avoid confusion about the sign conventions in the commutators.","section":"Eq. (64) and App. D"},{"comment":"In the Ω^{-2} coefficient, the expression '8π^4/37' appears arithmetically surprising; it may be a typographical artifact. Please check whether the intended denominator is 729 (or another value) and correct it.","section":"Eq. (D21)"}],"recommendation":"major_revision","confidential_remarks":"The reader's stress-test verification of the high-frequency coefficient (1/126) and my own independent check of the low-frequency coefficient (1/100) give me considerable confidence that the physics is correct. The requested revisions are therefore not changes of substance but rather completion of the manuscript: the low-frequency eigenvector calculation needs to be shown, and the van Vleck import should be accompanied by a brief justification or careful formal statement. The paper is within the scope of cond-mat.stat-mech and should be publishable after these repairs."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nThe central results hold up. For the three-step protocol, the high-frequency current is 1/126·(2π/Ω) and the low-frequency one is 1/100·(Ω/2π), with sign reversal under sequence reversal. I checked the high-frequency constant by direct computation from the effective generator, and I re-derived the low-frequency 1/100 via first-order perturbation theory on the dominant eigenmodes. Both match. That puts this paper well ahead of the typical driven-Markov-process paper, where the asymptotics are often just numerics plus a guess.\n\nWhat is new: the non-perturbative treatment of stepwise, finite-amplitude driving in a counting-field Floquet formalism. The van Vleck expansion for discrete protocols with Hurwitz zeta sums (Eq. 64) and the geometric-phase low-frequency formula adapted to stepwise switching are not in the earlier continuous-driving literature. The paper also does something honest: the J_dyn=0 choice is disclosed up front, and no constants are fitted to the numerics. The effective-rate interpretation in the high-frequency limit is a bonus that makes the result usable.\n\nNow the soft spots. The van Vleck expansion is imported from Hermitian quantum Floquet theory and applied to a non-Hermitian piecewise-constant generator without an error bound. I think this is a minor gap, for the reason in the stress-test note: for finite-dimensional bounded W(t) the one-period propagator is analytic in T near zero, so the expansion is a convergent Taylor expansion, not merely a formal series. But the paper does not make that argument, and a referee should ask for it. The low-frequency coefficient 1/100 is quoted without the eigenvector computation; that should be shown in an appendix. No code or data are included, which slowed verification. These are all addressable.\n\nThe citation practice is fine. Previous stochastic pumping work is credited properly, and the self-citations are background for quantum Floquet tools. The biological motivation is light but not oversold.\n\nBottom line: this deserves a serious referee. I'd accept it for review. With the omitted derivations supplied and a sentence or two about the non-Hermitian expansion, it would be a solid contribution to the driven stochastic kinetics literature. I'd cite it.","headline":"The Floquet current formulas in this paper are correct as far as I can verify; the only real gaps are rigor and reproducibility, not substance.","tokens_in":24472,"tokens_out":3163,"would_cite":true,"duration_ms":27127,"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":"Stepwise periodic switching of reaction rates in a Michaelis–Menten enzyme produces a net product current: (1/126)(2π/Ω) at high frequency and (1/100)(Ω/2π) at low frequency, with the sign set by the cycling direction of the rate steps.","keywords":["Floquet theory","counting field","stochastic processes","Markov generators","Michaelis–Menten kinetics","geometric phase","van Vleck expansion","discrete driving"],"falsifier":"Numerically integrate the two-state master equation for the three-step protocol γ_ac over many periods at Ω = 10π and measure the long-time slope of N_P(t); if it does not converge to (1/126)(2π/Ω) = 1/630 within the stated O(Ω⁻²) error, the high-frequency formula is wrong. At Ω = 0.2π, the low-frequency prediction is (1/100)(Ω/2π) = 1/1000, and both slopes should reverse sign when the step order is reversed to γ_c.","tokens_in":23520,"feed_emoji":"🧪","tokens_out":7403,"duration_ms":68727,"temperature":0.7,"pith_summary":"This paper sets out a Floquet theory for classical stochastic chemical reactions whose rate constants are switched periodically in time, formulated through a counting field. It claims that even without any net bias in the averaged rates, a long-time product current appears, controlled by the cyclic ordering of the driving steps and by the driving frequency Ω. At high frequency the current decays as 1/Ω and can be reinterpreted as a static enzyme with renormalized Michaelis–Menten rates; at low frequency the current grows linearly with Ω and arises from a geometric phase. The authors test the framework on a three-step, abruptly driven enzymatic model inspired by cAMP production and obtain analytical formulas that match direct numerical integration.","feed_headline":"Enzyme pacing creates a current that flips with cycle order","feed_subtitle":"Fast switching renormalizes the reaction rates; slow switching pumps geometrically. Both asymptotes match direct numerics.","key_machinery":"The central machinery is the counting-field-dependent Markov generator W(χ;t), whose one-period propagator is decomposed by Floquet theory into an effective Floquet generator W_eff(χ) and a kicked state |Ψ(0;t)⟩; the current is obtained from the χ-derivative of W_eff evaluated in that state. At high frequency, W_eff is constructed by a van Vleck expansion in powers of 1/Ω, with the commutators of Fourier components summed into closed forms involving the Hurwitz zeta function. At low frequency, repeated projection onto the instantaneous dominant eigenmode yields a geometric phase φ(χ), whose χ-derivative gives a current linear in Ω. The stepwise protocol is an ordered sequence γ = {W_0, …, W_","core_discovery":"The paper's central claim is that the long-time current in a periodically driven Markov process is controlled by a counting-field Floquet generator: J∞(Ω) = ⟨S|Ĵ_eff(0)|ψ0⟩, the derivative of the effective generator with respect to the counting field, evaluated in the kicked state. For the three-step protocol with generators (1,1,1,1) → (2,1,1,1) → (1,1,1,2), the authors derive explicit asymptotics: J∞(Ω) = (1/126)(2π/Ω) + O(Ω⁻²) in the high-frequency regime and J∞(Ω) = (1/100)(Ω/2π) in the low-frequency regime, and they show that reversing the cyclic order γ_ac → γ_c flips the sign. In the high-frequency limit the driven reaction is equivalent to a static enzyme with renormalized rates; in","pith_inferences":["A direct extension left implicit in the paper: higher derivatives of the cumulant generating function at χ = 0 should yield the full counting statistics — diffusion coefficient, skewness, and beyond — for product counts under periodic driving, so the same Floquet machinery could describe noise and rare-event statistics.","Because the N-step protocol has a continuous-driving limit as N → ∞, the derived high- and low-frequency formulas should reproduce earlier smooth-driving results in that limit; comparing the coefficients as N grows would test whether the asymptotic structure is universal across driving shapes.","The effective-rate picture at high Ω suggests a practical control scheme: by choosing Ω, one can tune the renormalized Michaelis–Menten parameters continuously, which could allow optogenetically switched enzymes to be programmed as tunable rate elements in a biochemical circuit.","The robustness of the geometric current to dwell-time redistribution implies that even irregular gating of an enzyme — where step intervals are not equal — would pump at the same per-period rate as long as each interval reaches quasi-steady state; this may be relevant to in vivo G-protein-driven cAMP signaling, though the paper does not make that claim."],"forward_implications":["Periodic switching alone, with no bias in the time-averaged rates, generates a net product current whose direction is set by the cyclic order of the switching steps.","In the high-frequency limit, the driven enzyme behaves like a static enzyme with renormalized reaction rates; for the three-step example the effective rate k1 increases and k_-2 decreases by an amount proportional to 1/Ω.","The long-time current scales as 1/Ω for fast switching and as Ω for slow switching, so the same protocol exhibits opposite frequency dependencies in the two regimes, with a crossover in between.","Reversing the driving sequence γ_ac to γ_c reverses the sign of the current in both the high- and low-frequency regimes.","The geometric part of the low-frequency current is independent of how the total period is split among the steps, as long as each step is long enough for relaxation; only the dynamical part depends on the dwell-time weights.","The current can vanish when averaged rates are balanced, yet appear purely from the ordering of the rate steps, establishing a stochastic-pumping mechanism under strong, abrupt driving."],"fun_headline_variants":["Periodic enzyme switching flips current with cycle order","Floquet-driven enzymes: high-frequency renormalizes rates","Enzyme current from Floquet driving: sign depends on order","Stepwise enzyme protocol yields long-time product current","Floquet theory for enzymatic reactions: current and statistics"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The load-bearing premise is that the van Vleck high-frequency expansion, imported from quantum Floquet theory, remains a valid asymptotic series for the non-Hermitian, abruptly switched Markov generator; the paper offers numerical agreement rather than a proven error bound.","fun_headline_variants_meta":{"raw":{"variants":["Periodic enzyme switching flips current with cycle order","Floquet-driven enzymes: high-frequency renormalizes rates","Enzyme current from Floquet driving: sign depends on order","Stepwise enzyme protocol yields long-time product current","Floquet theory for enzymatic reactions: current and statistics"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000553,"raw_usage":{"total_tokens":2537,"prompt_tokens":872,"completion_tokens":1665,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":616,"completion_tokens_details":{"reasoning_tokens":1584}},"tokens_in":616,"tokens_out":1665,"duration_ms":10847,"temperature":1.0,"reasoning_tokens":1584,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-01T19:09:05.941747+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Numerically integrate the two-state master equation for the three-step protocol γ_ac over many periods at Ω = 10π and measure the long-time slope of N_P(t); if it does not converge to (1/126)(2π/Ω) = 1/630 within the stated O(Ω⁻²) error, the high-frequency formula is wrong. At Ω = 0.2π, the low-frequency prediction is (1/100)(Ω/2π) = 1/1000, and both slopes should reverse sign when the step order is reversed to γ_c.","supporting_citations":[],"review_version":1}