{"id":"68f54d86-4387-469d-ae2f-4a3302b6effc","arxiv_id":"2509.08019","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":8,"one_line_summary":"A cellulose-peptide nano-paper selectively stores hydrophilic and lipophilic drugs in different nanoscale compartments; release and dose control were not tested.","lead":"Researchers created a nano-paper from cellulose and peptide hydrogel that stores water-loving and fat-loving drugs in separate nanometer-scale compartments, as shown by X-ray scattering. The material could enable future low-dose, site-specific drug delivery, but this study only shows storage, not release or dosing.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"No direct evidence localizes drugs in claimed CMC/P compartments; scattering-only inference leaves selective-storage claim underdetermined.","rationale":"The reader's weakest assumption identifies the same load-bearing gap: the mapping from scattering changes to drug location. I agree with that assessment. The paper has genuine strengths—time-resolved in-situ GISAXS/GIMAXS, kinetic differences between drugs, and FTIR showing peptide secondary-structure stability—but none of these directly prove that drug molecules reside in the intended CMC vs P nanodomains. The claim that TNP can be customized for all drugs, with controlled low-dose storage and site-specific release, rests on this compartment assignment. Since the SI (fitting details, Tables S3/S4) was not in the reviewed text, even the indirect structural evidence cannot be fully audited. The concern is not that the authors are wrong; it is that the presented data are compatible with several alternative interpretations, and the paper's central conclusion requires distinguishing among them. A direct chemical-mapping experiment (or a functional release measurement) would settle it. Therefore the reader's conditional verdict is appropriate; no change is needed.","tokens_in":13991,"tokens_out":4426,"duration_ms":52396,"concrete_test":"Prepare TNP films loaded with CM (sulfur), CQP (chlorine), and RD (phosphorus), cryo-section them, and acquire nanoscale chemical maps (STEM-EDX or synchrotron X-ray fluorescence) overlaid with a CMC/peptide marker (e.g., O/C/N ratios or labeled components). If the drug-specific element signals are not spatially enriched in the claimed compartments (S in CMC-rich regions; Cl/P in peptide-rich regions) and instead appear as surface layers or separate aggregates, the selective-storage interpretation is falsified. A release assay into PBS with HPLC-MS quantification would independently test whether the stored drug is functionally available.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim—that TNP selectively stores hydrophilic drugs in CMC domains and lipophilic drugs in peptide domains, making it an 'ideal carrier' for all drugs—requires that drug molecules are actually inside those molecular environments, molecularly dispersed. The presented evidence is entirely indirect: GISAXS/GIMAXS intensities change upon drug spraying, and fitted CMC1/CMC2 radii and the peptide fractal radius/correlation length shift in a logD7.4-dependent manner (Fig. 2C, Fig. 3G, Fig. 4). Equivalent scattering changes could arise from drug adsorbed on film surfaces, drug-rich crystalline/amorphous aggregates, drying-induced matrix reorganization, or film-thickness/roughness effects, none of which require selective internal compartment storage. The fitting details and tabulated values are in SI Tables S3/S4, which were not part of the reviewed text, so model uniqueness and fit quality cannot be checked. No chemical mapping of drug-specific elements and no release/dissolution measurement is reported. Without such functional/locational evidence, the strong 'ideal carrier for all drugs' conclusion is underdetermined.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper introduces a 'therapeutic nano-paper' (TNP) made of interwoven carboxymethylated cellulose (CMC) and peptide hydrogel (P) domains, and claims that this architecture can selectively store hydrophilic drugs in CMC domains and lipophilic drugs in the peptide mesh. The authors use in-situ, time-resolved GISAXS/GIMAXS during spray deposition of several drugs (CM, CQP, RD, NM) to extract domain radii (R_CMC1, R_CMC2, R_P), correlation lengths, and kinetic intensity traces. They correlate these structural parameters with externally tabulated logD7.4 and protein-binding values, and supplement the X-ray work with FTIR and DFT-computed molecular radii. The conclusion is that TNP is an 'ideal carrier platform' for all drug types, enabling low-dose controlled and site-specific release.","tokens_in":14288,"tokens_out":3549,"duration_ms":42584,"significance":"If the central claims were fully substantiated, the TNP concept could be a useful platform for personalized, low-dose drug delivery, and the combination of in-situ scattering with external lipophilicity data is attractive. The manuscript has genuine strengths: the GISAXS/GIMAXS experiments are time-resolved and in-situ, stated errors are provided for many fitted parameters, FTIR serves as a secondary structural probe, and the DFT radii and logD7.4/protein-binding data are independent of the scattering model. These features reduce (though do not eliminate) concerns about circularity. However, the paper's headline conclusions—selective compartment storage, autocatalytic embedding kinetics, and universal carrier suitability—are considerably broader than what the current evidence can support.","major_comments":[{"comment":"The central claim that hydrophilic drugs are stored in CMC domains and lipophilic drugs in the peptide mesh rests entirely on changes in fitted scattering radii (R_CMC1, R_CMC2, R_P). No direct chemical mapping, element-specific spectroscopy, or release/dissolution experiment verifies that the drugs are molecularly dispersed inside the intended compartments. Equivalent scattering changes could arise from surface adsorption, drug-rich crystalline/amorphous aggregates, drying-induced matrix reorganization, or film-thickness/roughness effects. The statement after Fig. 2C that 'The big changes in the GISAXS and GIMAXS data prove a significant embedding of the drugs into the TNP' overstates what scattering alone can prove. The authors should either add direct localization and release/loading measurements, or substantially temper the 'ideal carrier platform for all type of drugs' conclusion.","section":"Results and Discussion, Figs. 2C and 3G; Conclusions, final paragraph"},{"comment":"The 'autocatalytic' kinetic interpretation is not supported by the data as presented. The analysis shows normalized intensity decreases/increases of the CMC2 and P regions, but no kinetic model, rate law, or concentration-dependence test is given. The conclusion that 'the autocatalytic embedding of CM in TNP is unequivocally supported' is unjustified; a classical first-order or simple two-step process would produce similar normalized intensity traces. The authors should fit competing kinetic models with meaningful uncertainties or remove the autocatalytic claim.","section":"Fig. 5 and following bullet list; Conclusion, first paragraph"},{"comment":"The correlations between structural parameters and logD7.4/protein binding are based on at most four drugs (CM, CQP, RD, NM). No correlation coefficients, confidence intervals, or goodness-of-fit measures are reported. Some fitted values overlap within their stated errors (e.g., R_CMC1 = 200 ± 15 nm for CQP vs 190 ± 20 nm for RD), so the claimed monotonic trend in Fig. 4A is not robust. Given that the 'generalizable' conclusion depends on these correlations, a larger drug set and statistical treatment are needed, or the claims must be limited to the studied compounds.","section":"Fig. 4 and accompanying text"},{"comment":"The structural interpretation depends on the two-step cylinder/fractal fitting scheme, but the fitting details are relegated to SI Tables S3/S4, which were not available for review. Model uniqueness, parameter correlations, and fit quality cannot be checked. Since the domain-selective storage story relies on these fitted radii, the authors should provide the full fitting protocol, including parameter uncertainties and model comparison, either in the main text or in a reviewable supplement. The brief description of 'Int0' and 'Int1' in Fig. 3 is also insufficient for reproducibility.","section":"Results and Discussion, two-step analysis and SI Tables S3/S4"}],"minor_comments":[{"comment":"The caption says 'RM' but the text and context indicate 'RD' (Remdesivir). Please correct.","section":"Fig. 3F caption"},{"comment":"The sentence 'This relocation of the P-assembled structures to other areas within the TNP implies that the disordered arms of the P-mesh loosen their network between the fibres and compactly arrange themselves with the drugs around the fibres.' appears twice verbatim. Remove the duplicate.","section":"Bullet list after Fig. 5"},{"comment":"Define all drug abbreviations (CM, CQP, RD, NM) and give the full peptide sequence in the main text or a clear appendix footnote; the current 'foodnote' reference is incomplete.","section":"Throughout"},{"comment":"The phrase 'site-specific drug release on a beyond-nanomolar scale' and 'low-dose-controlled consumption' imply functional release data that are not presented. Please rephrase as potential or future work unless release measurements are added.","section":"Abstract and Conclusions"},{"comment":"The terms 'radii of gyration', 'fractal radius R_P', and 'aggregate radii R_CMC1/R_CMC2' are used somewhat interchangeably. Clarify the exact definitions and which quantity is plotted in each panel.","section":"Fig. 4 and text"}],"recommendation":"major_revision","confidential_remarks":"The manuscript contains interesting in-situ scattering data and an original material concept, but the published version overreaches: the selective-storage, dose-control, and release claims go beyond the evidence. I would not reject outright because the X-ray/FTIR dataset may be salvageable with substantial revision—either by adding direct localization/loading/release evidence or by carefully narrowing the conclusions. Please ensure the SI (Tables S3/S4) is included in any resubmission, as the fitting details are essential for evaluating the central claim."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nThe paper is a materials-science proof-of-principle: spray-deposited carboxymethyl cellulose / peptide hydrogel films are characterized with in-situ GISAXS/GIMAXS while drugs of different lipophilicity are added. The fitted radii of the cellulose and peptide domains shift with logD7.4 in a systematic way, and that is genuinely new. Best parts: the fits carry reported errors, the FTIR control shows the peptide secondary structure survives drug loading, and the external logD7.4/protein-binding values are an independent axis for the correlations.\n\nThe soft spot is exactly what the stress-test note says: there is no direct evidence that a drug molecule sits inside the CMC or P domain. The assignment comes from fitting model radii and correlation lengths, then interpreting those changes as selective compartmental uptake. Surface adsorption, drug aggregates, drying artifacts, or thickness changes could produce similar scattering shifts. No release data, no chemical mapping, no measured loading dose. The 'autocatalytic' kinetics are asserted from a couple of intensity curves, and that language is stronger than the evidence. The conclusion that TNP is an 'ideal carrier for all type of drugs' is a leap from a handful of compounds.\n\nEven with all that, I would send this to reviewers. The structural observation is plausible, the method is careful, and a good referee can push the authors to add direct localization evidence or at least temper the claims. The SI was not part of the reviewed text; the fitting details and model uniqueness need scrutiny before the domain assignment is accepted. If it holds up, this is a useful platform paper; if not, the scattering observations are still a contribution, just a more modest one.\n\nReading-group value is moderate. I would not cite it in my own near-term work, but it is the kind of paper that gets cited as a platform method if the SI checks out.\n\nRecommendation: send for peer review, require the SI, and ask the reviewers to verify the model uniqueness and the drug-localization interpretation.","headline":"A promising materials proof-of-principle that overreaches on therapeutic claims; the structural observations are worth a careful referee, but the selective-storage story is underdetermined.","tokens_in":14880,"tokens_out":2556,"would_cite":false,"duration_ms":28847,"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 its therapeutic nano-paper—a nanoscale weave of cellulose and peptide hydrogel—stores hydrophilic drugs in the cellulose domains, lipophilic drugs in the peptide mesh, and adapts its structure around each drug for low-","keywords":["therapeutic nano-paper","drug delivery","nanocellulose","peptide hydrogel","GISAXS/GIMAXS","lipophilicity","low-dose drug storage","in-situ X-ray scattering"],"falsifier":"A cross-sectional chemical map of a drug-loaded TNP film (for example Raman or element-specific microscopy) plus a sequential release assay that extracts the cellulose and peptide domains separately would settle where the drugs actually reside. If the hydrophilic drug were found outside the cellulose domains, or the lipophilic drug outside the peptide mesh, the domain-selective storage claim would collapse.","tokens_in":13917,"feed_emoji":"💊","tokens_out":11885,"duration_ms":111660,"temperature":0.7,"pith_summary":"This paper introduces 'therapeutic nano-paper' (TNP), a film built from alternating nanoscale layers of carboxy-methylated cellulose and a peptide hydrogel, and claims it can hold hydrophilic and lipophilic drugs simultaneously, each in the domain that matches its solubility. Using in-situ, time-resolved X-ray scattering while drug solutions are sprayed onto the growing film, the authors find that the most hydrophilic drug studied (camostat mesylate) swells the cellulose fiber domains, while progressively more lipophilic drugs (chloroquine phosphate, remdesivir) progressively swell the fractal peptide mesh. These structural responses track the drugs' published lipophilicity values (logD7.4 and protein binding), and the film's formation kinetics differ by drug, with hydrophilic loading triggering an autocatalytic reorganization of both networks and lipophilic loading enriching the peptide mesh. If correct, a single carrier material could be matched to whichever drug a treatment requires and deliver it in small, precisely controlled doses—relevant for potent drugs whose overdose risk demands low-dose control.","feed_headline":"Nano-paper stores water- and oil-loving drugs in separate pockets","feed_subtitle":"Cellulose and peptide meshes each take up the drug type that matches their solubility, enabling low-dose delivery.","key_machinery":"The central object is therapeutic nano-paper (TNP): a nano-stacked, interwoven template of carboxy-methylated cellulose (CMC) fibers and a peptide hydrogel (P) whose long chains, including hydrophobic glycine-lysine-phenylalanine-glycine (GLFG) stretches among hydrophilic residues, form a micellar fractal mesh. This adjacent pairing of a hydrophilic cellulose phase and a lipophilic peptide phase is what gives drug molecules of complementary solubility a matching domain. The argument rests on two scattering observables from grazing-incidence X-ray data: the cellulose fiber-aggregate radii (approximately 150 nm and 50 nm unloaded) from the small-angle regime, and the peptide mesh's 'fractal ra","core_discovery":"The paper's central claim is that its therapeutic nano-paper adapts around each drug: hydrophilic drugs are stored in the cellulose domains, lipophilic drugs in the peptide mesh. Hydrophilic camostat mesylate swells the large cellulose fibers from 150 to 280 nm radius (medium fibers from 50 to 200 nm); lipophilic chloroquine phosphate and remdesivir leave cellulose nearly unchanged and instead swell the peptide fractal radius from 0.47 to 0.74 and 1.22 nm. Spray-fabrication kinetics are drug-specific: hydrophilic loading autocatalytically reorganizes both networks; lipophilic loading enriches the peptide mesh. FTIR shows the peptide beta-sheet structure survives loading. The authors conclude","pith_inferences":["A direct test the paper does not make: measuring release profiles from the two compartments separately, or chemically mapping drug position inside the film, would confirm the compartment assignment that the scattering analysis infers.","The authors' conclusion that TNP can be customized for 'all type of drugs' goes beyond the studied set of model drugs; it is a generalization from a proof-of-principle, not a demonstrated universal.","If domain-selective uptake holds, a natural extension is co-loading two drugs of opposite solubility into the two subphases at once—a combination therapy in a single patch—which the paper motivates but does not demonstrate.","The logD7.4 and protein-binding correlations suggest a quantitative prediction rule: given a drug's tabulated lipophilicity, one could anticipate whether it will swell the cellulose domain or the peptide mesh before running the scattering experiment."],"forward_implications":["Hydrophilic and lipophilic drugs can be loaded into the same carrier without chemical modification, each partitioning into the domain matching its solubility.","Dose can be controlled at the nanoscale by tuning drug concentration and domain sizes during spray fabrication, addressing the overdose risk of potent drugs.","The peptide mesh's fractal radius responds to molecular size as well as lipophilicity (0.47 to 1.22 nm across the studied drugs), implying the carrier can be tuned for a range of drug sizes, not just solubility classes.","Because structural response correlates with published logD7.4 and protein-binding values, a drug's expected storage compartment could be predicted from tables before any experiment.","Each drug has a characteristic embedding half-time (20-62 s in the studied spray cycles), so fabrication parameters can be chosen per drug to avoid incomplete loading or precipitation."],"supporting_citations":[{"why":"supplies the hydrodynamic nanocellulose filament assembly on which the CMC subphase of the TNP is built","marker":"[15]"},{"why":"provides the peptide-hydrogel meshwork whose fractal network structure the P-domain analysis extends","marker":"[18]"},{"why":"supplies the FTIR amide-band assignment used to conclude the peptide beta-sheet structure survives drug loading","marker":"[27]"},{"why":"outlines the fractal-analysis scheme the paper applies to the molecular-scale GIMAXS intensities","marker":"[31]"},{"why":"provides the scattering-model software used in the fractal refinement of the peptide network","marker":"[34]"},{"why":"establishes the in-situ spray-deposition scattering methodology used for the kinetic measurements","marker":"[24]"},{"why":"is the database source of the logD7.4 lipophilicity values against which the structural radii are correlated","marker":"[37]"},{"why":"supplies the drug-solubility and protein-binding definitions used as the second lipophilicity axis","marker":"[38]"}],"fun_headline_variants":["Nano-paper sorts drugs by solubility into separate meshes","Cellulose-peptide nano-paper packs each drug type in its matching pocket","Drug-matching nano-paper stores hydrophilic and lipophilic separately","Nano-paper with twin meshes stores drugs by water or oil affinity","Selective nano-paper pockets hold drugs by solubility match"],"cache_read_input_tokens":2688,"weakest_assumption_plain":"The load-bearing assumption is that the scattering-visible swelling of a nanoscale domain really means the drug molecules are stored inside that domain; the paper presents no direct chemical mapping, spectroscopic localization, or release data that would rule out drugs sitting on the film surface or in separate crystalline phases.","fun_headline_variants_meta":{"raw":{"variants":["Nano-paper sorts drugs by solubility into separate meshes","Cellulose-peptide nano-paper packs each drug type in its matching pocket","Drug-matching nano-paper stores hydrophilic and lipophilic separately","Nano-paper with twin meshes stores drugs by water or oil affinity","Selective nano-paper pockets hold drugs by solubility match"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000528,"raw_usage":{"total_tokens":2347,"prompt_tokens":672,"completion_tokens":1675,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":416,"completion_tokens_details":{"reasoning_tokens":1584}},"tokens_in":416,"tokens_out":1675,"duration_ms":14421,"temperature":1.0,"reasoning_tokens":1584,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-04T22:04:59.094146+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A cross-sectional chemical map of a drug-loaded TNP film (for example Raman or element-specific microscopy) plus a sequential release assay that extracts the cellulose and peptide domains separately would settle where the drugs actually reside. If the hydrophilic drug were found outside the cellulose domains, or the lipophilic drug outside the peptide mesh, the domain-selective storage claim would collapse.","supporting_citations":[{"cited_title":"Håkansson, A","cited_arxiv_id":null,"evidence_quote":"supplies the hydrodynamic nanocellulose filament assembly on which the CMC subphase of the TNP is built"},{"cited_title":"Petri, S","cited_arxiv_id":null,"evidence_quote":"provides the peptide-hydrogel meshwork whose fractal network structure the P-domain analysis extends"},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"supplies the FTIR amide-band assignment used to conclude the peptide beta-sheet structure survives drug loading"},{"cited_title":"Quevedo, M","cited_arxiv_id":null,"evidence_quote":"outlines the fractal-analysis scheme the paper applies to the molecular-scale GIMAXS intensities"},{"cited_title":"Doucet et al","cited_arxiv_id":null,"evidence_quote":"provides the scattering-model software used in the fractal refinement of the peptide network"},{"cited_title":"Zhang, G","cited_arxiv_id":null,"evidence_quote":"establishes the in-situ spray-deposition scattering methodology used for the kinetic measurements"},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"is the database source of the logD7.4 lipophilicity values against which the structural radii are correlated"},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"supplies the drug-solubility and protein-binding definitions used as the second lipophilicity axis"}],"review_version":1}