{"id":"45d2cbab-acc7-4eb7-bf95-2cae9eae2a4b","arxiv_id":"2508.15091","paper_version":1,"verdict":"UNVERDICTED","confidence":"LOW","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"First direct AFM visualization of sequence-dependent poloidal orientation in DNA minicircles, with coarse-grained simulations matching the observed phase shifts.","lead":"Using atomic force microscopy, researchers showed that short circular DNA molecules adopt a sequence-dependent preferred twist, the poloidal orientation, which had been predicted by theory but never directly seen. The work introduces a single-molecule labeling method that maps which face of the DNA ring points inward.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Insert-sequence confound: the 20-nt biotin-bearing insert, not the 85-nt core, may set the inside/outside phase; no control visible in abstract or figure captions.","rationale":"The reader's weakest_assumption points exactly to the 20-nt insert as a potential confound. My reading of the abstract and Figure S1 supports that this is the most load-bearing issue: the experimental readout is physically attached to the insert; the core sequence is the independent variable, but the label position is on a different sequence element. Since the full text is corrupted and no controls are visible in the abstract/captions, the central claim cannot be verified from the available material. Therefore the reader's UNVERDICTED disposition is appropriate, and our concern does not change that verdict, but it sharpens the specific test that would resolve it.","tokens_in":2499,"tokens_out":3632,"duration_ms":42637,"concrete_test":"Use the authors' coarse-grained model to simulate each 85-nt core with (i) the actual 20-nt insert, (ii) a neutral/random 20-nt insert of matched length, and (iii) no insert (bare core). Compare the predicted inside/outside phase vs insertion position. If the phase difference between 601 and A-tract cores changes by more than the experimental phase uncertainty when the insert is replaced or removed, the insert is not neutral and the AFM readout is confounded. Complementarily, simulate a random 85-nt core with the same insert; a nonzero phase would indicate insert-dominated orientation.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim is that the phase of inside/outside NeutrAvidin positions reports the poloidal orientation of the two 85-nt core sequences. But the position readout is on a 20-nt insert containing a biotin-modified dT, and Figure S1 shows all constructs are core + insert at one of seven helical positions. If the insert itself has an intrinsic curvature (or a bend induced by biotin/NeutrAvidin), the observed phase could reflect the insert's orientation in the minicircle rather than the core's poloidal orientation. This is not a remote possibility: the insert is a sequence-specific 20-mer, and biotin-dT modification can perturb local DNA structure. Because the two cores are different sequences, the local stacking at the core–insert junctions differs, so even a single fixed insert can produce a differential phase shift between the 601 and A-tract series that has nothing to do with the cores' isolated bend preferences. The abstract and legible captions do not state whether the insert sequence is identical across constructs, whether control inserts with different bend propensities were tested, or whether the coarse-grained simulations included the insert explicitly and with what parameters. Without that information, the experiment does not cleanly separate the insert's own orientation from the core's poloidal orientation, and the headline conclusion is not yet supported.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper introduces an AFM-based single-molecule assay to visualize the poloidal orientation (the inside/outside face) of short DNA minicircles. A biotin within a 20-nt insert is placed at different positions along an 85-nt core, and the location of NeutrAvidin relative to the minicircle loop is measured. For two core sequences (the 601 half and an A-tract sequence), the inside/outside probability varies with insert position with different phases, which the authors interpret as distinct preferred poloidal orientations. Coarse-grained simulations are reported to show narrow poloidal distributions with different mean orientations, consistent with the AFM data. The paper claims the first experimental confirmation of sequence-dependent poloidal orientation in DNA minicircles.","tokens_in":2790,"tokens_out":4817,"duration_ms":57155,"significance":"If the interpretation is correct, this is the first direct experimental observation of a theoretically predicted but previously unseen phenomenon: sequence-dependent poloidal orientation in short DNA minicircles. The AFM readout is externally grounded: the position of a protein label does not presuppose the orientation it is used to infer, so the study is not circular. The agreement with coarse-grained simulations, if genuine, would be a valuable cross-validation. However, the version of the manuscript under review is not verifiable: the main text is largely unreadable due to character corruption, and the legible parts leave a potentially load-bearing confound unaddressed.","major_comments":[{"comment":"The central readout is the position of NeutrAvidin attached to a biotin-modified dT within a 20-nt insert. Figure S1 shows that every construct is an 85-nt core plus this insert, but the insert sequence is not given and no control for its own bend preference is described. If the insert has an intrinsic curvature or the biotin-dT perturbs local stacking, the inside/outside phase may be set by the insert rather than by the core; because the two cores differ, the core-insert junction also differs, so a differential phase between the 601 and A-tract series need not reflect the cores' poloidal orientations. The authors should state whether the insert sequence is identical in all constructs, include controls with alternate insert sequences, and/or show that simulations omitting the insert reproduce the same phase shift.","section":"Figure S1 / Abstract"},{"comment":"The supplied main text is not readable: large portions consist of garbled characters, so the experimental protocol, the image-analysis procedure, the definition of the poloidal angle, and the simulation details cannot be checked. This is not a mere presentation issue; it makes verification of the central claim impossible in the version provided. A complete, readable manuscript with full Methods is required.","section":"Full text / Methods"},{"comment":"The abstract states that coarse-grained simulations 'revealed' a narrow distribution and different mean orientation for each sequence, but no simulation parameters or fitting procedure are legible. If the sequence-dependent stiffness parameters were adjusted to reproduce the AFM phase shift, then the agreement is not independent confirmation. The authors must report whether the parameters are taken from prior work or fitted here, and must quantify the comparison (e.g., predicted vs measured phase for all seven insert positions).","section":"Abstract / Coarse-grained simulations"},{"comment":"The histograms in Figure S4 are presented without sample sizes, fit functions, or statistical uncertainties. The claims of 'narrowly distributed' poloidal angles and of a significant phase difference between the 601 and A-tract series require a quantitative comparison (e.g., circular statistics, confidence intervals, or a permutation test). Without these, the quoted phase shift is not supported.","section":"Figure S4"}],"minor_comments":[{"comment":"The caption says the insert position 'changes from 0 to 12', which is more than one helical turn (10.5 bp/turn). Please clarify whether the positions are along the 85-nt core or phasing relative to the helical repeat, and specify the exact total length of each minicircle (85+20 bp).","section":"Figure S4 caption"},{"comment":"Figure S1 refers to Table S1 for the full sequence list, but Table S1 is not legible in the supplied version. Please ensure all supplementary tables are included and readable.","section":"Figure S1 / Table S1"},{"comment":"The abstract uses 'short (<150 bp)' but the construct appears to be 105 bp. Stating the exact length and confirming the ligation product is a monomeric circle (e.g., by gel or AFM contour-length control) would strengthen the description.","section":"Abstract"}],"recommendation":"major_revision","confidential_remarks":"The version of the manuscript I received has a corrupted main text, which may be a submission/PDF conversion artifact rather than the authors' fault, but it must be fixed before any further review. The most serious scientific issue is the unaddressed 20-nt insert confound: the label carrier itself may set the measured phase. Because the conclusion rests on the phase difference between two core sequences, a control or analysis that removes the insert contribution is essential. I recommend major revision rather than rejection because the conceptual approach is sound and the issue appears addressable with additional experiments or re-analysis."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The premise is solid: theory has predicted sequence-dependent poloidal orientation for short DNA minicircles, and there's been no direct experimental test. The design is genuinely clever—biotin at different positions along a helical turn, AFM readout of inside/outside—and the abstract's claim is crisp: two sequences show different phases, and coarse-grained simulations match. If the data hold up, this is a real first.\n\nThe soft spot is load-bearing: the readout is on a 20-nt insert, not on the core sequence that is supposed to set the poloidal orientation. The insert is a sequence-specific piece of DNA with a biotin-dT, sitting at one of seven positions. If that insert has any intrinsic bend—or bends differently at the two core junctions—the observed inside/outside phase could report the insert's local orientation rather than the core's poloidal orientation. The abstract and the visible captions do not say whether the insert sequence is identical across constructs, whether the simulations include it, or whether there was any control for insert bending. Without that, the headline conclusion does not cleanly separate the label carrier from the thing being measured.\n\nThe full text is corrupted in the arXiv posting, so I can't check whether the authors addressed this in the main text. That's a practical problem for review but not a scientific one, assuming the authors have the real text. If they can show that the insert is explicitly modeled in the simulation and that the phase pattern is insensitive to its own bend, the paper is likely solid. As is, it's an unverified claim with a plausible confound.\n\nWho's this for? Single-molecule biophysicists and DNA mechanics people. The technique itself could be useful even if the specific conclusion needs more work. If the full text is intact, send it to review, but ask the referee to scrutinize the insert handling. I wouldn't cite it yet.","headline":"Clever first experimental attack on a real prediction, but the label-carrying insert is a plausible confound that the available text doesn't rule out.","tokens_in":3257,"tokens_out":2437,"would_cite":false,"duration_ms":27956,"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":"This paper reports experimental confirmation of sequence-dependent poloidal orientation in DNA minicircles, using AFM to see whether a biotin label lands inside or outside the circle.","keywords":["DNA minicircle","poloidal orientation","atomic force microscopy","single-molecule biophysics","sequence-dependent DNA bending","601 nucleosome positioning sequence","A-tracts","coarse-grained simulation"],"falsifier":"Repeat the same two core sequences with a set of biotin inserts whose sequences are permuted or reversed at the same positions: if the inside/outside phase follows the insert sequence rather than the core, the claim that the two cores have distinct poloidal orientations collapses. A simpler check is to place the same insert sequence in different registers along the core and see whether the phase shifts track the insert's own bend direction.","tokens_in":2390,"feed_emoji":"🧬","tokens_out":5054,"duration_ms":57167,"temperature":0.7,"pith_summary":"Short DNA circles are predicted to bend unevenly because sequence-dependent energetics define a preferred side of the duplex to face inward—the poloidal orientation. The paper introduces an AFM-based single-molecule readout for that orientation: each minicircle carries a single biotin at one of seven positions along one helical turn, and bound NeutrAvidin marks whether that position sits inside or outside the circle. Applying this to two sequences predicted to have strong preferences—one derived from the 601 nucleosome positioning sequence, one from six in-phase A-tracts—the authors observe opposite phase shifts in the inside/outside position, and coarse-grained simulations reproduce narrowly distributed, sequence-specific orientations. If correct, these results give direct experimental evidence that poloidal orientation is real and sequence-dependent, supporting the view that nonuniform bending energetics shape the dynamics of circular DNA.","feed_headline":"DNA minicircles show sequence-dependent inside-out orientation","feed_subtitle":"Two test sequences flip their labels to opposite faces of the circle, confirming predicted poloidal orientation.","key_machinery":"The central object is poloidal orientation: the angle specifying which side of the DNA double helix faces the center of a minicircle versus facing outward. The readout machinery is a biotin ruler—seven constructs with a single biotin placed at successive positions along one helical turn—imaged by AFM through bound NeutrAvidin, which converts the unobservable bending-face angle into a binary inside/outside measurement. Coarse-grained simulations of the same sequences supply the predicted angular distributions that the AFM phases are compared against.","core_discovery":"The paper's central claim is that a short DNA minicircle adopts a preferred poloidal orientation—a sequence-determined choice of which face of the double helix bends toward the center of the circle. To test this, the authors made minicircles from an 85-nt core taken from either the left half of the 601 nucleosome positioning sequence or from six in-phase A-tracts, inserting a 20-nt biotin-tagged segment at seven positions spaced along one helical turn. AFM images scored whether the NeutrAvidin label appeared inside or outside the circle. For the two cores, the inside/outside position shifted with different phase as the insert moved around the helix, indicating distinct mean poloidal orientat","pith_inferences":["If the insert-position confound is controlled, the same ladder could map poloidal orientation across many sequence motifs, turning AFM into a readout for the direction, not just the magnitude, of sequence-dependent DNA bending.","Because the orientation determines which bases face inward, protein-binding sites on the inner face of a small circle could be sterically occluded, which would matter for DNA looping and chromatin studies.","The phase-shift method could be extended to detect orientation changes induced by protein binding, salt, or supercoiling, since such perturbations should rotate the phase of the inside/outside signal.","The close match with simulation suggests the models could predict the orientation of a new sequence before synthesis, enabling design of minicircles with labels or reactive groups deliberately placed on the inside or outside surface."],"forward_implications":["Poloidal orientation is measurable and sequence-specific, so a short DNA circle's shape is defined not only by its curvature but by which face of the helix bends inward.","Two known bending-relevant motifs, the 601 nucleosome positioning sequence and in-phase A-tracts, have opposite preferred orientations, meaning sequence identity controls how a minicircle presents its surface.","Coarse-grained simulations that include sequence-dependent bending can predict the orientation, giving a computational handle for designing circular DNA with intended inside/outside features.","The biotin-position ladder provides a general AFM assay for mapping the bending landscape of arbitrary roughly 85-bp sequences.","The result connects sequence-dependent DNA mechanics to topologically constrained DNA behavior such as looping, where the inward-facing side of a circle is the side most relevant for protein access."],"supporting_citations":[],"fun_headline_variants":["DNA minicircles orient inside-out by sequence, AFM confirms","Bend direction of DNA circles depends on sequence, visualized","AFM reveals DNA minicircles adopt sequence-determined folds","DNA minicircles pick a face to bend inward, sequence decides","First images show DNA minicircles flip labels by sequence"],"cache_read_input_tokens":2688,"weakest_assumption_plain":"The load-bearing premise is that the measured inside/outside position of the NeutrAvidin label reports the poloidal orientation of the 85-nucleotide core sequence, rather than being dominated by the bend preference of the 20-nucleotide biotin-tagged insert that carries the label.","fun_headline_variants_meta":{"raw":{"variants":["DNA minicircles orient inside-out by sequence, AFM confirms","Bend direction of DNA circles depends on sequence, visualized","AFM reveals DNA minicircles adopt sequence-determined folds","DNA minicircles pick a face to bend inward, sequence decides","First images show DNA minicircles flip labels by sequence"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000188,"raw_usage":{"total_tokens":1179,"prompt_tokens":763,"completion_tokens":416,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":507,"completion_tokens_details":{"reasoning_tokens":326}},"tokens_in":507,"tokens_out":416,"duration_ms":5630,"temperature":1.0,"reasoning_tokens":326,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-05T18:06:55.245854+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Repeat the same two core sequences with a set of biotin inserts whose sequences are permuted or reversed at the same positions: if the inside/outside phase follows the insert sequence rather than the core, the claim that the two cores have distinct poloidal orientations collapses. A simpler check is to place the same insert sequence in different registers along the core and see whether the phase shifts track the insert's own bend direction.","supporting_citations":[],"review_version":1}