{"id":"a0a81247-c097-4eba-914a-be1ff44f7fcc","arxiv_id":"2504.19766","paper_version":1,"verdict":"REJECT","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"A combined IR, CD, and MD study reports that the Omicron S1 spike subunit is more hydrophilic, has lower beta-sheet content, and retains an open conformation compared to Alpha and Gamma.","lead":"This paper compares the structure of the SARS-CoV-2 spike S1 subunit from the Alpha, Gamma, and Omicron variants using infrared and circular dichroism spectroscopy plus molecular dynamics simulations. It reports that Omicron S1 is more hydrophilic, has less beta-sheet, and stays in an open conformation, which may relate to its higher infectivity.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The only-Omicron-open-state claim rests on one MD replica per starting state and a post-hoc state-averaging choice; without replicate simulations and matched state selection, the central conformational difference is unsupported.","rationale":"The paper's headline contribution is the joint claim that Omicron S1 has lower beta-sheet content and is the only variant retaining an open state, which is then linked to ACE2 affinity and transmissibility. The most load-bearing condition for that claim is that the MD simulations represent the equilibrium conformational behavior of each variant. That condition is not met by the reported design: exactly one 600 ns trajectory per starting state, with no replicates and no assessment of run-to-run variability. Large domain rearrangements in spike proteins are slow, and a single trajectory cannot distinguish a sequence-encoded preference from stochastic relaxation or force-field bias. The problem is visible in the paper's own language: Section 3.3 describes Omicron 'maintaining' Rg, but the Discussion converts this into 'the only one retaining an open state' and 'stronger stability in the RBD-up configuration.' The MD secondary-structure comparison in Table 4 is also biased: Alpha and Gamma are represented only by their closed-state trajectories, while Omicron is represented by an average of open and closed states. This post-hoc choice was made after observing the Rg behavior and makes the MD-derived beta-sheet difference non-comparable. The reported experimental beta-sheet values are also not cleanly 'beyond error bars': CD values overlap (37 ± 7 vs 38 ± 7), IR values overlap at one SD (35 ± 3 vs 39 ± 2), and the MD comparison is state-biased. A replicate protocol with matched state selection would directly test whether the unique open-state retention is real. Until then, the central claim is under-supported, consistent with the reader's REJECT verdict.","tokens_in":23368,"tokens_out":6944,"duration_ms":76370,"concrete_test":"Run at least 5 independent 600 ns MD replicas per variant and per starting state (Alpha/Gamma/Omicron × open/closed), with randomized initial velocities and at least two distinct ColabFold model seeds. Report per-replica final NTD-RBD distance and the fraction of the last 100 ns spent with Rg > 4 nm. If any Alpha or Gamma replica remains open, or any Omicron replica closes, the single-trajectory uniqueness claim is not supported. Recalculate MD secondary-structure comparisons using identical state-selection criteria for all three variants.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim that Omicron S1 uniquely retains an open/RBD-up conformation is supported only by a single 600 ns MD replica per starting state (Section 3.3, Table 3, Figure 5). There is no replicate, no occupancy statistic, and no check that the Alpha/Gamma closures are not simply initial-model relaxation or force-field drift. The problem is compounded by Table 4: Alpha and Gamma MD secondary-structure values are taken from closed-state trajectories only, while Omicron values are averaged over open and closed states, so the lower Omicron beta-sheet content in MD is confounded by post-hoc state selection. In addition, the simulations use unglycosylated monomeric ColabFold models, whereas the experimental proteins are glycosylated; the monomeric NTD-RBD distance defining open/closed is not demonstrated to correspond to the RBD-up/down equilibrium of the trimeric spike. Unless the open-state retention is reproducible across independent replicas and consistent under matched state-selection and glycosylation conditions, the variant-specific conformational claim is not established.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript reports a comparative biophysical study of the monomeric S1 subunits of SARS-CoV-2 Alpha, Gamma, and Omicron spike proteins, combining ATR-IR spectroscopy, circular dichroism (CD) spectroscopy, ColabFold structure prediction, 600 ns molecular dynamics simulations, and Protein-sol surface calculations. The authors estimate secondary-structure percentages from each technique, interpret the amide I red shift of Omicron as evidence of increased hydrophilicity, and conclude that Omicron S1 has lower beta-sheet content than Alpha and Gamma, a more hydrophilic surface, and uniquely retains an open (RBD-up-like) conformation in MD simulations, which they connect to higher ACE2 affinity and transmissibility. The claimed novelty is the systematic combination of these experimental and computational techniques on three variants of concern.","tokens_in":23556,"tokens_out":5406,"duration_ms":57219,"significance":"If the claims were established, the paper would provide a useful multi-technique comparison of variant spike proteins and would strengthen the case for IR/CD spectroscopy as rapid structural screens for emerging variants. The authors should be credited for the multi-technique design, the explicit reporting of mutation tables and sequences, and the use of commercially relevant S1 proteins. However, the central claims are not supported by the evidence as presented: the reported error bars for beta-sheet content overlap, the MD secondary-structure comparison is confounded by inconsistent state selection, and the open-state retention claim rests on a single trajectory per starting state. The dataset may still be a useful contribution after substantial revision, but the current conclusions outrun the statistical and methodological basis.","major_comments":[{"comment":"The statement that Omicron has a lower beta-sheet content \"which represents a significant difference with respect to Alpha and Gamma beyond the error bars\" is contradicted by the paper's own uncertainties. In Table 1, CD-derived beta-sheet values are Alpha 38 ± 7 and Omicron 37 ± 7; in Table 2, IR-derived values are Alpha 39 ± 2 and Omicron 35 ± 3, a difference that is approximately 1.1 combined standard deviations; in Table 4, MD-derived values are Alpha 37 ± 3 and Omicron 35 ± 2. The weighted averages shown in Figure 8 therefore almost certainly have overlapping error bars. The claim of significance should either be supported by an explicit statistical test (e.g., a t-test or bootstrap on the per-deposition estimates) or be downgraded to a trend.","section":"Section 4, Tables 1, 2, 4, Figure 8"},{"comment":"The MD secondary-structure comparison is confounded by inconsistent state selection. For Alpha and Gamma, the percentages in Table 4 are taken from closed-state trajectories only, while for Omicron the values are explicitly \"an average between open and closed states.\" Because open and closed states differ in radius of gyration, solvent exposure, and likely secondary-structure composition, this averaging can artificially lower the Omicron beta-sheet content relative to the other variants. State-specific values should be reported separately for all variants, and any ensemble average should be justified with Boltzmann weights or occupancy statistics rather than a simple arithmetic mean.","section":"Section 3.3, Table 4"},{"comment":"The claim that Omicron \"is the only one retaining an open state when starting from an open state\" is based on a single 600 ns MD replica per starting state, as stated in the Methods for a single replica of each model. With no replicate simulations, no occupancy statistics, and no check that the Alpha and Gamma closures are not initial-model relaxation or force-field drift, the conformational difference is not statistically established. This concern is compounded by two additional issues: the simulations use unglycosylated monomeric ColabFold models, whereas the experimental proteins are glycosylated, and the NTD-RBD center-of-mass distance used to define open/closed states is not demonstrated to correspond to the RBD-up/RBD-down equilibrium of the trimeric spike. Replicate trajectories, matched glycosylation status, and validation of the open/closed metric are needed before this central claim can be accepted.","section":"Section 2.4, Section 3.3, Table 3, Figure 5"}],"minor_comments":[{"comment":"Table 1 reports results from the CLUSTR algorithm, but Section 2.3 states that CONTINLL, CDSSTR, and SELCON3 were employed; please reconcile this inconsistency.","section":"Section 2.3, Table 1"},{"comment":"The optical path length of the CD cuvette is given as 0.01 nm, which is physically implausible; this should presumably be 0.01 cm.","section":"Section 2.3"},{"comment":"In the Omicron row, the random coil entry is printed as \"32 5\" instead of \"32 ± 5.\"","section":"Table 1"},{"comment":"The force field is referred to as CHARMM22/CMAP in the text and CHARMM22* in the discussion; please use a consistent nomenclature throughout.","section":"Section 2.4"},{"comment":"The NPP surface ratio results are presented only as qualitative color maps; quantitative values or distributions for the RBD and NTD regions would strengthen the hydrophilicity comparison.","section":"Section 3.4, Figures 6 and 7"},{"comment":"The statement that \"noticeable differences\" occur in the CD spectra is not supported by any statistical comparison of replicate spectra; showing error envelopes or replicate curves would make this assessment more rigorous.","section":"Section 3.1"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is best treated as a major revision rather than a rejection, because the experimental dataset is potentially useful and the methodological improvements needed are identifiable. However, if the authors cannot provide replicate MD simulations, matched state selection, and a defensible significance test for the beta-sheet comparison, the central claims should be substantially downgraded or the paper should not be published."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The paper gives you a new comparative dataset: ATR-IR, CD, and 600-ns MD on monomeric S1 from Alpha, Gamma, and Omicron at pH 7.4. That dataset is genuinely useful, and the combination of three methods on matched commercial proteins is not something I've seen before. The amide I redshift and the component changes in Omicron are described carefully, and the NPP surface calculations add a plausible structural rationale.\n\nBut the central quantitative claim does not hold up. The Discussion says Omicron has a significantly lower beta-sheet content, beyond the error bars, compared to Alpha and Gamma. Look at their own numbers: CD gives 38±7, 39±6, 37±7; IR gives 39±2, 40±3, 35±3; MD gives 37±3, 39±2, 35±2. In every case the differences are within about one standard deviation. The IR difference is the largest, but with n=5 depositions and those error bars, calling it 'beyond the error bars' is not justified. No significance test is reported. The conclusion should be softened to a trend, not a claim.\n\nThe MD open-state claim is also under-supported. The authors run a single 600-ns replica per starting state. Alpha and Gamma close; Omicron stays open. With no replicates, you cannot distinguish mutation-driven differences from stochastic fluctuation or initial-model relaxation. The paper's own logic makes this worse: for the MD secondary-structure comparison (Table 4), Alpha and Gamma are reported from closed trajectories only, while Omicron is an average of open and closed. That is a post-hoc selection that biases the beta-sheet comparison in the direction of the claimed result. And the simulations use unglycosylated ColabFold models of monomeric S1, whereas the experimental proteins are glycosylated and in the trimeric spike context; the NTD-RBD distance used to define open/closed is not shown to correspond to the RBD up/down equilibrium.\n\nWhat's good: the experimental side is careful. Five independent depositions, water subtraction, reproducibility check, CDPro with multiple algorithms and basis sets, error propagation on the Gaussian fits. The observed spectral differences — amide I redshift, changes in the beta-sheet component distribution — are real and worth reporting, even if the interpretation is speculative.\n\nWho is this for? People working on spectroscopic biosensing of SARS-CoV-2 variants, or on comparative structural biology of the spike. It's not a definitive answer on Omicron's RBD-up stability, but it's a solid descriptive dataset that generates a hypothesis.\n\nMy recommendation: send it to peer review. The authors need to fix the statistics, add replicate MD runs (or at least remove the 'beyond the error bars' language), and address the state-selection inconsistency. With those changes, this could be a useful methods-and-dataset paper.","headline":"Useful comparative spectroscopic dataset, but the headline claims about Omicron's beta-sheet content and open-state stability are not supported by the paper's own uncertainties.","tokens_in":24128,"tokens_out":2593,"would_cite":false,"duration_ms":25593,"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 Omicron's S1 spike subunit shows less beta-sheet content, a more hydrophilic surface, and a stronger tendency to stay in the open, RBD-up conformation than Alpha or Gamma, helping to explain its higher ACE2 affinity.","keywords":["SARS-CoV-2 variants","spike glycoprotein S1","secondary structure","amide I infrared spectroscopy","circular dichroism","molecular dynamics","RBD-up conformation","hydrophilicity"],"falsifier":"A concrete test would be to run multiple independent 600 ns molecular dynamics replicas of Alpha, Gamma, and Omicron S1 from open starting states and count how often each closes; if Alpha or Gamma remain open as often as Omicron, or if Omicron frequently closes, the conformational claim fails. An experimental cross-check is small-angle X-ray scattering of the three isolated S1 subunits in buffer, which would show a larger radius of gyration for Omicron than for Alpha and Gamma if the open state is truly retained.","tokens_in":1943,"feed_emoji":"🦠","tokens_out":4604,"duration_ms":112863,"temperature":0.7,"pith_summary":"This paper aims to establish that the S1 spike subunit of the Omicron variant of SARS-CoV-2 is measurably different from the Alpha and Gamma variants, and that the differences are visible both spectroscopically and in simulation. It reports a lower beta-sheet fraction for Omicron (about 35% versus 38–40% for Alpha and Gamma), a redshifted amide I band consistent with stronger hydrogen bonding to water, more hydrophilic surface patches on the receptor-binding and N-terminal domains, and a molecular dynamics trajectory in which Omicron stays open when started open while Alpha and Gamma close. If these differences are real, they would help explain Omicron's stronger binding to ACE2 and its enhanced transmissibility, and they would show that infrared and circular dichroism spectroscopy can serve as fast, non-destructive fingerprints of variant-specific protein conformation.","feed_headline":"Omicron spike protein is measurably more open and hydrophilic","feed_subtitle":"Alpha and Gamma S1 subunits close up in simulations; Omicron stays open, matching its stronger ACE2 binding.","key_machinery":"The investigative engine is a three-way comparison of the monomeric S1 subunit (the part of the spike containing the N-terminal domain and receptor-binding domain), with open versus closed state tracked by the NTD–RBD center-of-mass distance (threshold about 5 nm) and by the radius of gyration $R_g$. The spectroscopic workhorse is the amide I band (1590–1720 cm$^{-1}$), deconvoluted into $\\beta$-sheet, $\\alpha$-helix, random-coil, and $\\beta$-turn components, with the $\\beta$-sheet signal split into the characteristic low-frequency $\\nu_\\perp$ and high-frequency $\\nu_\\parallel$ modes; CD spectra deconvoluted with three algorithms over six reference sets give an independent secondary-structure estimate. The computational complement is 600 ns molecular dynamics with free-energy surfaces in $R_g$/RMSD space, plus non-polar-to-polar surface ratio maps that locate hydrophilic patches on the RBD and NTD. The key interpretive link is the amide I redshift: lower vibrational frequencies are read as stronger hydrogen bonding to water, which ties together the lower $\\beta$-sheet content, the more hydrophilic surfaces, and the tendency to remain open.","core_discovery":"Combining attenuated total reflection infrared spectroscopy, circular dichroism, and molecular dynamics, the authors claim that the three variants share broadly similar secondary structure—about 27–28% random coil, 24–27% $\\beta$-turn, 8–9% $\\alpha$-helix—but Omicron stands apart in $\\beta$-sheet content: 35±3% by IR, versus 39±2% for Alpha and 40±3% for Gamma, a gap the paper calls significant beyond error bars. The amide I maximum shifts from 1650 cm$^{-1}$ (Alpha) to 1648 cm$^{-1}$ (Gamma) to 1647 cm$^{-1}$ (Omicron), with all Omicron components slightly redshifted, which is interpreted as stronger C=O hydrogen bonding with water and therefore greater hydrophilicity. Surface polarity maps show larger hydrophilic areas on Omicron's RBD and NTD than on Alpha's, and MD simulations show that the open-state Omicron model keeps a radius of gyration near 4.3 nm over 600 ns, whereas Alpha and Gamma open-state models collapse to closed configurations. The paper identifies the open state with the RBD-up conformation of the full spike and concludes that Omicron's S1 favors RBD-up, matching its documented higher ACE2 affinity and infectivity.","pith_inferences":["Because the open-versus-closed conclusion rests on a single 600 ns trajectory per starting state, a natural next test is to run many independent replicas; if Alpha and Gamma sometimes stay open as often as Omicron, the mutation-specific interpretation would not survive.","The simulation models lack the sugar modifications present on the measured proteins, so the paper's link between simulation and spectroscopy implicitly assumes glycans do not change the open/closed balance; this could be tested by simulating a glycosylated S1 or by comparing deglycosylated samples spectroscopically.","If the amide I redshift really tracks hydrophilicity, the same infrared protocol could be applied to newer variants to see whether the $\\beta$-sheet fraction continues to drop as the spike accumulates mutations.","The S1 monomer is studied without the S2 fusion machinery or the viral membrane; whether the monomer's open preference persists in the full prefusion spike is a separate, testable question."],"forward_implications":["A spectroscopic lab could in principle distinguish Omicron S1 from Alpha or Gamma by the combination of a redshifted amide I peak and reduced $\\beta$-sheet fraction, without needing antibodies or sequencing.","If Omicron S1 genuinely favors the open, RBD-up state, that would provide a structural reason for its enhanced ACE2 binding and faster spread relative to earlier variants.","The trend in the amide I maximum (1650, 1648, 1647 cm$^{-1}$ for Alpha, Gamma, Omicron) tracks the mutation count (7, 10, 31), suggesting that accumulated mutations progressively alter the protein's hydration and conformational ensemble.","The more hydrophilic character of Omicron's RBD and NTD implies that water-mediated contacts, not just direct amino-acid contacts, contribute to its stronger receptor interaction."],"supporting_citations":[{"why":"Earlier IR measurements of coronavirus spike proteins establish the measurement protocol and provide the wild-type amide I reference that this study compares against.","marker":"[58]"},{"why":"Previous amide I deconvolution of SARS-CoV-2 RBD and spike proteins supplies the component assignment and side-chain subtraction method used here.","marker":"[59]"},{"why":"Reference for amino-acid side-chain infrared absorption, used to subtract their contribution before secondary-structure quantification.","marker":"[37]"},{"why":"Defines the CD secondary-structure estimation method that yields the CD-derived percentages for the three variants.","marker":"[66]"},{"why":"Provides the CD reference basis sets used to deconvolute the variant spectra.","marker":"[65]"},{"why":"Structure-prediction method used to build the open and closed S1 models that seed the molecular dynamics simulations.","marker":"[57]"},{"why":"Establishes the correspondence between S1 open/closed configurations and the RBD up/down states of the full spike, which the discussion relies on.","marker":"[93]"},{"why":"Shows Omicron mutations stabilize the up-RBD conformation, an independent result the authors use to support their interpretation.","marker":"[101]"},{"why":"Single-molecule measurements of variant-ACE2 interactions cited as evidence of Omicron's stronger receptor binding.","marker":"[16]"}],"fun_headline_variants":["Omicron spike S1 stays open and wetter than Alpha or Gamma","IR and MD show Omicron spike is more open and hydrophilic","Omicron's S1 beta-sheet content drops, matching open RBD-up state","Spike spectroscopy: Omicron is more hydrophilic and open","Omicron spike's structural fingerprint: less beta-sheet, more solvent exposure"],"cache_read_input_tokens":26368,"weakest_assumption_plain":"The load-bearing premise is that a single 600 ns molecular dynamics run per starting state, on predicted models that lack the sugar modifications, faithfully represents how the glycosylated proteins behave in solution; if that sampling or model choice is unrepresentative, the claim that Omicron alone retains the open state—and the explanation built on it—does not follow.","fun_headline_variants_meta":{"raw":{"variants":["Omicron spike S1 stays open and wetter than Alpha or Gamma","IR and MD show Omicron spike is more open and hydrophilic","Omicron's S1 beta-sheet content drops, matching open RBD-up state","Spike spectroscopy: Omicron is more hydrophilic and open","Omicron spike's structural fingerprint: less beta-sheet, more solvent exposure"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000723,"raw_usage":{"total_tokens":3278,"prompt_tokens":1015,"completion_tokens":2263,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":631,"completion_tokens_details":{"reasoning_tokens":2168}},"tokens_in":631,"tokens_out":2263,"duration_ms":16488,"temperature":1.0,"reasoning_tokens":2168,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-16T05:43:45.643936+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A concrete test would be to run multiple independent 600 ns molecular dynamics replicas of Alpha, Gamma, and Omicron S1 from open starting states and count how often each closes; if Alpha or Gamma remain open as often as Omicron, or if Omicron frequently closes, the conformational claim fails. An experimental cross-check is small-angle X-ray scattering of the three isolated S1 subunits in buffer, which would show a larger radius of gyration for Omicron than for Alpha and Gamma if the open state is truly retained.","supporting_citations":[{"cited_title":"D’Arco, M","cited_arxiv_id":null,"evidence_quote":"Earlier IR measurements of coronavirus spike proteins establish the measurement protocol and provide the wild-type amide I reference that this study compares against."},{"cited_title":"Mancini, S","cited_arxiv_id":null,"evidence_quote":"Previous amide I deconvolution of SARS-CoV-2 RBD and spike proteins supplies the component assignment and side-chain subtraction method used here."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Reference for amino-acid side-chain infrared absorption, used to subtract their contribution before secondary-structure quantification."},{"cited_title":"Sreerama, R.W","cited_arxiv_id":null,"evidence_quote":"Defines the CD secondary-structure estimation method that yields the CD-derived percentages for the three variants."},{"cited_title":"Sreerama, R.W","cited_arxiv_id":null,"evidence_quote":"Provides the CD reference basis sets used to deconvolute the variant spectra."},{"cited_title":"Mirdita, K","cited_arxiv_id":null,"evidence_quote":"Structure-prediction method used to build the open and closed S1 models that seed the molecular dynamics simulations."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Establishes the correspondence between S1 open/closed configurations and the RBD up/down states of the full spike, which the discussion relies on."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Shows Omicron mutations stabilize the up-RBD conformation, an independent result the authors use to support their interpretation."}],"review_version":1}