{"id":"5016e139-60e0-4c6c-af35-bfa691dc1745","arxiv_id":"2412.12127","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":5,"one_line_summary":"A second-quantized electron-phonon model produces a sharp co-resonance peak in the electron current cross-spectrum for EcoRI and its cognate DNA sequence, which disappears for randomized target sites.","lead":"This paper models how electrons moving along a DNA chain and a restriction enzyme could resonate at matching frequencies, and it finds a sharp resonance only when the DNA carries the enzyme's target sequence. The result is a step toward explaining how proteins find their DNA targets at a distance, though no physical force is actually computed.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The co-resonance claim rests entirely on the EIIP empirical scale; if those values are not transferable to hydrated biomolecules, the sequence specificity is an artifact of the input table, and no independent validation is provided.","rationale":"The reader's weakest assumption captures the key fragility: the EIIP table is the sole carrier of sequence information. I agree with that assessment, and I do not find a different concern that is more load-bearing. The central claim would be true only if EIIP values are meaningful physical descriptors under biological conditions. The proposed test settles this by varying the input scale while holding the model fixed. Because the paper's own cross-checks (RRM, mutation tests) use the same EIIP input, they cannot independently validate the table. A null permutation test is particularly decisive: if the peak is an artifact of the particular numbers, a shuffle of the same numbers should destroy it for the cognate pair; if the dynamics are robust, the peak should remain for sequence-specific arrangements only. No change to the reader's CONDITIONAL verdict is needed; this is exactly the condition that must be met.","tokens_in":15513,"tokens_out":8470,"duration_ms":83885,"concrete_test":"Recompute the cross-frequency spectra of Figs. 2-7 with the identical protocol, but replace the EIIP site energies by a physically grounded alternative, e.g., B3LYP/6-31G* vertical ionization potentials (or electron affinities) of the 4 nucleotides and 20 amino acids in a PCM water model, or by a null model in which the EIIP values are randomly permuted among the 66 nucleotide sites and 276 amino acid sites while preserving the multiset. If the sharp co-resonance for the cognate site persists across these different assignments, the result is robust to the EIIP table; if it disappears, the central claim is contingent on that specific empirical scale and the conditional verdict is not satisfied.","verdict_should_be":"UNCHANGED","load_bearing_attack":"All sequence-dependent physics enters through the EIIP tables (Refs. [31,32]): the electron tunneling amplitudes J_n are computed from E0 and the barriers E_{n+1} via Eqs. (15)-(16), and the electron-phonon coupling is set to chi_n = (E_{n+1}-E_n)/a. The paper gives no argument that these 1972 pseudopotential values are transferable to hydrated, flexible biomolecules at physiological pH and ionic strength; it does not compare them with ab initio site energies or with any experimental electron-affinity or ionization data. The agreement with the RRM peak (Ref. [34]) is not independent evidence, since RRM uses the same EIIP values, so both calculations share the same potential systematic error. The Appendix A mutation tests are similarly generated from the same table and therefore cannot validate the table. Hence the sharp co-resonance peak for CTTAAG and its broadening for randomized or mutated sequences, the central claim, is entirely conditional on the empirical validity of the EIIP assignments. If those assignments are arbitrary or environment-dependent, the sequence-specific terahertz resonance is a property of a lookup table, not of DNA-protein electrodynamics.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proposes a second-quantized Davydov/Holstein-Fr\"ohlich model of electron transport along DNA and protein chains, with site-dependent tunneling and electron-phonon couplings derived from the EIIP tables of Refs. [31,32]. The authors numerically integrate the resulting TDVP equations of motion for a 66-nucleotide DNA fragment containing the EcoRI recognition site and for the 276-amino-acid EcoRI enzyme, compute the electron currents, and form their cross-frequency spectrum. They report a sharp co-resonance peak near 20-29 THz for the cognate CTTAAG sequence, which broadens or disappears for randomized and mutated recognition sequences, and they compare this with the RRM peak. Appendix A reports that single-point mutations in EcoRI that experimentally relax specificity retain the peak, whereas a mutation that reduces catalytic activity suppresses it. The paper closes with a speculative discussion of water-mediated and radiative mechanisms for long-range electrodynamic interactions.","tokens_in":15728,"tokens_out":6465,"duration_ms":63687,"significance":"If the central claim holds, the paper would provide a concrete, mechanistically explicit route from DNA/protein sequence to a sequence-specific terahertz resonance that could mediate long-range recognition, complementing the authors' earlier experimental work. The manuscript has real strengths: the equations of motion follow from a standard variational treatment, the numerical integration conserves energy to high accuracy, the Matlab scripts are deposited on Zenodo, and the Appendix A mutation tests are checked against independent experimental reports [46,47] that were not used to set the model parameters. However, the sequence-dependent physics enters entirely through the EIIP tables, and the reported spectra appear to be single thermal realizations. The significance of the paper is therefore conditional on demonstrating that the co-resonance is robust to EIIP uncertainty and to thermal initial-condition sampling.","major_comments":[{"comment":"The sequence-dependent content of the model is entirely supplied by the EIIP values in Tables I and II. The manuscript borrows these values from Refs. [31,32] but gives no argument that these pseudopotential values, originally derived for isolated atoms, remain valid for nucleotides and amino acids in hydrated, flexible biomolecules at physiological pH and ionic strength. Since J_n and chi_n are computed directly from E_n through Eqs. (15)-(16), any environmental or methodological dependence of the EIIP values propagates into the electron currents and hence into the co-resonance peak, which is the central claim. The agreement with the RRM peak in Figure 2(a) cannot resolve this concern, because RRM uses the same EIIP tables. I would like to see either a comparison with ab initio site energies or experimental electron-affinity/ionization data, or a sensitivity analysis in which the EIIP values are perturbed within a chemically plausible range and the survival of the co-resonance is quantified.","section":"III (Tables I-II) and Eqs. (15)-(16)"},{"comment":"The phonon initial conditions are drawn from zero-mean random values with the amplitudes of Eq. (28), but the reported spectra appear to correspond to a single realization. With no ensemble averaging or error bars, the sharpness of the co-resonance peak and its disappearance upon mutation could be a fluctuation of the thermal initial conditions. This concern is reinforced by the observation that the peak position shifts from about 20 THz in Figure 2(a) to about 29 THz in Figure 4(a) under different parameter sets; the manuscript does not report how many initial conditions were sampled or how stable the peak position and amplitude are. I request ensemble averaging over the random phonon initial conditions, with the mean spectrum and confidence intervals reported for the cognate and mutant cases.","section":"IV (Eq. (28) and Figs. 2-5)"},{"comment":"The claim of 'very good quantitative agreement' with the RRM peak [34] is not an independent check, because RRM is based on the same EIIP values (Refs. [24,25,32]) that generate the co-resonance in this paper. The quantitative comparison therefore largely shows that the present spectra inherit the RRM frequency content through the input tables. The phrase should be tempered, and the independent support for sequence sensitivity should rest primarily on the Appendix A mutation tests, which do provide a genuine comparison with experimental specificity data and should be presented as the main external validation.","section":"IV and Ref. [34]"}],"minor_comments":[{"comment":"The sentence 'This is in line with the attempt to understand whether intermolecular electrodynamic interactions are implicated...' appears twice, verbatim, in the introduction; one occurrence should be deleted.","section":"I (Introduction, p. 3-4)"},{"comment":"The text defining P_n has a typographical error: it reads 'Pn[b(t), q(t), p(t)' without a closing parenthesis, and Eq. (25) uses Pn with four arguments while the earlier definition has three; please harmonize the notation.","section":"Eqs. (24)-(25)"},{"comment":"The caption mixes dimensionless and dimensional parameters with inconsistent notation (e.g., Omega'_{1,n} versus Omega_{1,n}); please define every symbol once and use the same notation throughout.","section":"Figure 2 caption"},{"comment":"The y-axis scale and normalization of the cross-spectra are not defined, which makes it difficult for the reader to judge the sharpness or broadness of the peaks across panels; please specify how the cross-spectrum is normalized.","section":"IV (Figs. 2-5)"},{"comment":"References [22] and [38] are described as 'in peer review' and 'in submission'; please update their publication status or remove the status notes if they are no longer accurate.","section":"References [22] and [38]"}],"recommendation":"major_revision","confidential_remarks":"The main risk is that the central claim is conditioned on the empirical validity of the EIIP tables and on a single thermal realization. I believe this can be addressed within the scope of the manuscript by adding a sensitivity analysis with respect to the EIIP values and by ensemble-averaging the spectra. If the authors cannot provide such tests, the paper should be reframed as a demonstration of the model's internal coherence rather than as a quantitative prediction of a physical resonance. The Appendix A mutation comparison is the most convincing part of the paper and should be given more prominence."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The real content here is a numerical experiment: a Davydov/Holstein-Fröhlich model with site-dependent couplings, applied to EcoRI and its DNA target, showing a sharp cross-spectrum peak for the cognate CTTAAG sequence that broadens under single-base, double-base, and randomized mutations. That part is new, internally consistent, and the authors ship the Matlab code and check energy conservation. The Appendix A mutation tests are also a genuine plus: the promiscuous mutants retain a sharp peak and the catalytic mutant loses it, which matches published experiments not used to fit anything. If you accept the EIIP values as physical, this is a legitimate extension of the Resonant Recognition Model with a concrete dynamical mechanism.\n\nThe soft spot is exactly what the stress-test note says: all sequence-dependent physics enters through the EIIP tables from Refs. [31,32]. Tunneling amplitudes, electron-phonon coupling, everything. The paper gives no argument that these 1970s pseudopotential numbers survive contact with hydrated, flexible biomolecules, and no comparison to ab initio or experimental site energies. The agreement with RRM is not independent evidence, because RRM uses the same table. So the central claim—a sequence-specific terahertz resonance—is a property of the lookup table unless the table is independently validated. That is a load-bearing weakness, not a minor one.\n\nOther issues are lesser but real. The spectra appear to be single thermal realizations of the random initial phonon conditions; there are no error bars or ensemble averages. The co-resonance peak moves from about 20 THz in Figures 2–3 to 29 THz in Figures 4–5 under hand-chosen parameters, which undercuts the quantitative claim. And the title promises electrodynamic forces, but no force is ever computed; the authors explicitly defer that to future work, which the title and abstract overstate.\n\nThis is not a paper to reject out of hand. It is a coherent, reproducible model study of a real open problem, with honest acknowledgment that the biological relevance needs quantitative force estimates. It deserves a serious referee, but the referee should push hard on the EIIP transferability, demand ensemble averaging, and ask for a revised title that matches the actual claims. I would not cite it in its current form, but I would read a revised version.","headline":"A reproducible but conditional computational demonstration that sequence-specific co-resonance in DNA–EcoRI electron currents depends entirely on the empirical EIIP lookup table, whose biological validity is never independently tested.","tokens_in":16330,"tokens_out":1574,"would_cite":false,"duration_ms":16803,"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":"A second-quantized electron–phonon model predicts a sharp, sequence-specific terahertz co-resonance between EcoRI and its DNA target, lost when the recognition sequence is randomized.","keywords":["electrodynamic interactions","DNA–protein recognition","EcoRI restriction enzyme","electron–phonon coupling","Davydov model","Holstein–Fröhlich model","terahertz co-resonance","electron current spectra"],"falsifier":"Run the same simulation with the EIIP values randomly permuted among the four nucleotides or among the twenty amino acids while keeping the sequences fixed: if a sharp co-resonance peak survives the permutation, the peak is an artifact of the parameterization rather than of the sequence. On the experimental side, terahertz time-domain spectroscopy on hydrated EcoRI–DNA mixtures should show a narrow sequence-specific resonance near 20–29 THz for the cognate site that disappears for a randomized site; its absence would refute the predicted co-resonance.","tokens_in":15253,"feed_emoji":"🧬","tokens_out":7768,"duration_ms":68256,"temperature":0.7,"pith_summary":"This paper argues that selective DNA–protein recognition can be mediated by long-range electrodynamic forces, and it proposes a concrete dynamical mechanism: resonant electron currents. Using a second-quantized Hamiltonian of Davydov and Holstein-Fröhlich type, the authors derive electron–phonon dynamics for a DNA fragment and the EcoRI restriction enzyme, then compute the Fourier spectra of the electron currents on each molecule. The cross-spectrum shows a sharp co-resonance peak when the DNA contains EcoRI's cognate recognition sequence, while randomized or mutated sites replace the peak with a broad, noisy spectrum. Single-base changes degrade the peak progressively, and enzyme mutations known to relax or impair specificity behave accordingly. If correct, the model gives a sequence-specific, testable terahertz signature of DNA–protein recognition at a distance.","feed_headline":"EcoRI's electron currents co-resonate only with its DNA target","feed_subtitle":"A sharp peak near 20 THz appears only for the cognate site; single-base changes smear it, matching mutant data.","key_machinery":"The machinery is a one-dimensional, second-quantized electron–phonon Hamiltonian $\\hat{H} = \\hat{H}_{\\mathrm{el}} + \\hat{H}_{\\mathrm{ph}} + \\hat{H}_{\\mathrm{int}}$, with site-dependent tunnelling amplitudes $J_n$ and electron–phonon couplings $\\chi_n$. $J_n$ is computed as $E_0$ times a transmission coefficient across square barriers whose heights are the Electron–Ion Interaction Potential (EIIP) values of the nucleotide or amino acid at the next site, and $\\chi_n$ is set to $(E_{n+1} - E_n)/a$. A Davydov ansatz plus the time-dependent variational principle turns the operator equations into coupled classical-like equations for electronic amplitudes $C_n(t)$ and lattice displacements $\\beta_n(t)$; integrating these equations at room temperature yields the electron current along each chain. The central object is the cross Fourier spectrum $\\tilde{i}_1^*(\\nu)\\tilde{i}_2(\\nu)$ of the DNA and enzyme currents, whose sharp peak for the cognate sequence is the predicted selectivity signature.","core_discovery":"On the paper's own terms, the discovery is that the electron currents flowing along a DNA oligonucleotide and along the EcoRI enzyme, obtained from a site-dependent electron–phonon Hamiltonian, have strongly correlated time-domain Fourier spectra only when the DNA carries EcoRI's recognition sequence. The cross-correlation spectrum of the two currents exhibits a sharp co-resonance peak around 20 THz under one parameter set and around 29 THz under another, and this peak is replaced by a broad and noisy spectrum when the recognition sequence is randomized, when individual bases are exchanged with their complements, or when specificity-destroying mutations are introduced. The same calculation preserves the sharp peak for 'promiscuous' EcoRI mutants that retain relaxed binding. The authors propose this sequence-dependent co-resonance as evidence for selective electrodynamic interactions between DNA and proteins, complementing diffusive search mechanisms.","pith_inferences":["Inference: The peak position is not a fixed molecular constant: it shifts from about 20 THz to about 29 THz when spring constants and excitation energies change, so the model predicts that solvent conditions and hydration should tune the resonance frequency for a given sequence pair.","Inference: The same parameterization could be run for other restriction enzymes and transcription factors; a co-resonance at their cognate sites would support a general electrodynamic recognition mechanism, while failure would limit the mechanism to EcoRI-like systems.","Inference: Because all sequence dependence enters through the EIIP tables, permuting those values while keeping the sequences fixed is a numerical control the paper does not perform; without it, the sharp peak cannot be cleanly separated from the parameterization.","Inference: The direct current–current coupling uses a single electron and does not estimate force magnitudes; the paper's own water-mediated picture suggests the biologically relevant channel may be collective water polarization rather than direct electron-current attraction."],"forward_implications":["The cognate palindromic site yields a sharp cross-spectrum peak near 20 THz for one parameter set and near 29 THz for another, while randomized sites give broad, noisy spectra.","Single-base substitutions already broaden or reduce the peak, and double substitutions broaden it further, so the co-resonance is sequence-specific at single-base resolution.","Promiscuous EcoRI mutants that bind noncognate sites still show a sharp co-resonance peak, whereas a catalytically impaired mutant shows a sizeable decrease, matching experimental specificity data.","If the co-resonance operates in vivo, it provides a physical mechanism for long-range electrodynamic recognition that could accelerate encounters beyond Brownian diffusion.","The mechanism may generalize: any DNA-binding protein with a defined recognition sequence could exhibit a corresponding co-resonance with its target."],"supporting_citations":[{"why":"Supplies the EIIP tables for nucleotides and amino acids from which the site-dependent tunnelling amplitudes and electron–phonon couplings are computed; all sequence dependence enters through these values.","marker":"[31,32]"},{"why":"Provides the Davydov model, the amino-acid sound speed, and the sech-shaped initial electron wavefunction used in the simulations.","marker":"[27]"},{"why":"Earlier work by the authors on electron currents along a DNA chain that motivates computing current spectra for the DNA–enzyme pair.","marker":"[26]"},{"why":"Originates the Resonant Recognition Model and the use of EIIP sequences for digital signal processing of proteins and DNA.","marker":"[24,25]"},{"why":"Gives the RRM cross-correlation peak for EcoRI–DNA against which the computed ~20 THz co-resonance is compared.","marker":"[34]"},{"why":"Reports the experimental observation of long-distance electrodynamic intermolecular forces that motivates the search for a selective mechanism.","marker":"[21]"},{"why":"Provides the water-mediated electrodynamic interaction picture and the aromatic-network analysis used in Section V to connect the co-resonance to biological context.","marker":"[23]"},{"why":"Identifies promiscuous EcoRI mutants that bind noncognate sites; used in Appendix A to test that relaxed specificity keeps the co-resonance peak.","marker":"[46]"},{"why":"Shows the Asp91 to Asn mutation reduces EcoRI catalytic activity; used in Appendix A to test that impaired function lowers the co-resonance peak.","marker":"[47]"}],"fun_headline_variants":["Electron currents in DNA and EcoRI co-resonate only for target DNA","DNA-protein electron co-resonance: sequence-specific peak at THz","EcoRI's electron currents sync with its DNA target at THz frequencies","Sharp THz co-resonance between DNA and enzyme only when sequence matches","Selective electrodynamic grip: DNA-enzyme co-resonance depends on sequence"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the published electron–ion interaction energy values used for nucleotides and amino acids remain valid for hydrated molecules inside cells, because every sequence-dependent term in the model is built from those numbers.","fun_headline_variants_meta":{"raw":{"variants":["Electron currents in DNA and EcoRI co-resonate only for target DNA","DNA-protein electron co-resonance: sequence-specific peak at THz","EcoRI's electron currents sync with its DNA target at THz frequencies","Sharp THz co-resonance between DNA and enzyme only when sequence matches","Selective electrodynamic grip: DNA-enzyme co-resonance depends on sequence"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000886,"raw_usage":{"total_tokens":3847,"prompt_tokens":990,"completion_tokens":2857,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":606,"completion_tokens_details":{"reasoning_tokens":2756}},"tokens_in":606,"tokens_out":2857,"duration_ms":19002,"temperature":1.0,"reasoning_tokens":2756,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T21:54:13.530952+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Run the same simulation with the EIIP values randomly permuted among the four nucleotides or among the twenty amino acids while keeping the sequences fixed: if a sharp co-resonance peak survives the permutation, the peak is an artifact of the parameterization rather than of the sequence. On the experimental side, terahertz time-domain spectroscopy on hydrated EcoRI–DNA mixtures should show a narrow sequence-specific resonance near 20–29 THz for the cognate site that disappears for a randomized site; its absence would refute the predicted co-resonance.","supporting_citations":[{"cited_title":"Faraji, R","cited_arxiv_id":null,"evidence_quote":"Provides the Davydov model, the amino-acid sound speed, and the sech-shaped initial electron wavefunction used in the simulations."},{"cited_title":"Veljkovic, I","cited_arxiv_id":null,"evidence_quote":"Earlier work by the authors on electron currents along a DNA chain that motivates computing current spectra for the DNA–enzyme pair."},{"cited_title":"Cruzeiro, J","cited_arxiv_id":null,"evidence_quote":"Gives the RRM cross-correlation peak for EcoRI–DNA against which the computed ~20 THz co-resonance is compared."},{"cited_title":"Nardecchia, J","cited_arxiv_id":null,"evidence_quote":"Reports the experimental observation of long-distance electrodynamic intermolecular forces that motivates the search for a selective mechanism."},{"cited_title":"Examining the origins of observed terahertz modes from an optically pumped atomistic model protein in aqueous solution","cited_arxiv_id":"2210.14912","evidence_quote":"Provides the water-mediated electrodynamic interaction picture and the aromatic-network analysis used in Section V to connect the co-resonance to biological context."},{"cited_title":"Buchanan, The long reach of dipoles , Nature Phys","cited_arxiv_id":null,"evidence_quote":"Identifies promiscuous EcoRI mutants that bind noncognate sites; used in Appendix A to test that relaxed specificity keeps the co-resonance peak."},{"cited_title":"Sapienza, C.A","cited_arxiv_id":null,"evidence_quote":"Shows the Asp91 to Asn mutation reduces EcoRI catalytic activity; used in Appendix A to test that impaired function lowers the co-resonance peak."}],"review_version":1}