{"id":"a6101fde-8fd6-4f30-8cc9-e680ebf9ce73","arxiv_id":"2608.00425","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":4.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"The AFLOW prototype encyclopedia added 344 crystal structures, reaching 2,127 entries, along with database identifiers and tutorials.","lead":"The AFLOW library of crystallographic prototypes now contains 2,127 entries, 344 more than the prior release, with new ICSD and CCDC database cross-references and new tutorials. The paper documents the update and reviews research software that uses AFLOW prototype labels for structure identification and generation.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Appendix contains an internally inconsistent new prototype entry: Br2H12MgO12 listed under P42mc(105) carries label _cP108_205_, showing the new-entry list is not machine-validated.","rationale":"The reader's weakest assumption was that the 344 new prototype entries are accurate in space group, Wyckoff positions, and AFLOW label, and that no systematic labeling error exists. The appendix supplies a concrete counterexample: the new entry Br2H12MgO12 is printed under the tetragonal space group P42mc (No. 105) but its label encodes cubic space group Pa-3 (No. 205) with Pearson symbol cP. This is an internal inconsistency in the very evidence presented for the new entries, so the concern is not hypothetical. The paper deserves credit for making the prototype data publicly available on GitHub and aflow.org, and for documenting the software ecosystem that consumes AFLOW labels; the count of 2,127 prototypes may well be correct. However, because the paper explicitly omits any validation methodology or comparison against the source database entries, the discovered mismatch supports the reader's conditional verdict rather than a full acceptance. The test I propose is a mechanical audit that would settle whether this is an isolated typo or a symptom of a broader labeling problem; until such an audit is reported, the accuracy of the 344 additions remains an open condition.","tokens_in":70206,"tokens_out":5344,"duration_ms":48242,"concrete_test":"Write a script that parses Appendix A's section headers and each prototype label's Pearson symbol and space-group number, then flags every label whose space-group number or Pearson symbol does not match its section header. Run it over all 230 sections. For the flagged entries, starting with Br2H12MgO12, download the structure from the aflow.org repository, run AFLOW-XtalFinder or spglib to determine the actual space group and Wyckoff positions, and compare with both the printed label and the section header. If mismatches beyond the single visible one appear, the list of 344 new prototypes is not self-consistent and the paper's 'curated library' claim requires either correction or an explicit validation statement.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim is that the AFLOW library now contains 2,127 curated prototypes after adding 344 new entries, with the appendix serving as the evidence for these additions. In Appendix A, under the header 'P42mc(105)', the new entry (green bullet) 'Br 2H12MgO12: A2B12CD12_cP108_205_c_2d_a_2d-001 9A0M' has an AFLOW label whose space-group number is 205 (Pa-3, Pearson symbol cP), while the section header is space group 105 (P42mc, tetragonal, tP). These are different space groups and different Bravais lattices; this is not a typographical variant. Because the paper provides no automated cross-check, validation metrics, or comparison against source ICSD/CSD entries for the 344 additions, this visible mismatch is direct evidence that the published listing is not internally consistent. The AFLOW prototype label is the core artifact used by downstream software (XtalFinder, pymatgen, OpenKIM, NOMAD, robocrystallographer) to identify structures, so even a single mislabeled or mis-sorted entry compromises the claim of a reliable curated encyclopedia. The flaw may be in the paper's typesetting rather than the online database, but either way it demonstrates the absence of validation for the new-entry list.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript reports the fifth update of the AFLOW Library of Crystallographic Prototypes, claiming 344 new prototypes for a total of 2,127 entries. It also describes the addition of ICSD and CCDC deposition numbers on the web interface, new video and written crystallography tutorials, and a survey of research software that consumes AFLOW prototype labels or aflow.org data. The Appendix lists all prototypes ordered by space group, with new entries marked by green bullets.","tokens_in":70446,"tokens_out":4719,"duration_ms":45123,"significance":"If the data are accurate, this is a useful community resource: the count of 2,127 curated prototypes is central, the public GitHub repository makes the underlying data reproducible, and the survey of downstream tools such as XtalFinder, pymatgen, OpenKIM, NOMAD, Jmol, and ML-based generative models documents real adoption. The paper is a data descriptor rather than a hypothesis-testing study, so the absence of statistical error bars is not itself a problem; however, database papers stand or fall on the verifiability of the curated content, and the current appendix does not yet demonstrate that verifiability.","major_comments":[{"comment":"Under the header 'P42mc(105)', the new entry 'Br 2H12MgO12: A2B12CD12_cP108_205_c_2d_a_2d-001 9A0M' carries an AFLOW label whose Pearson symbol is cP and whose space-group number is 205 (Pa-3, cubic), while P42mc is space group 105 with tetragonal Pearson symbol tP. This is not a typographical variant of the same label: the two symbols denote different Bravais lattices and different space groups. Because the appendix is the evidence for the 344 new entries, this visible inconsistency demonstrates that the published new-entry list has not been machine-validated; at minimum the entry or the header must be corrected, and the list should be checked programmatically against the AFLOW prototype label conventions before publication.","section":"Appendix A (P42mc(105) entry)"},{"comment":"The paper states that 'All ICSD entries have a corresponding entry in the CCDC' and that both deposition numbers are therefore included. This is a substantive database claim, and no evidence is given; it is not obviously true because the ICSD and the CCDC CSD have different scopes, with the CSD primarily covering organic and metal-organic structures. If the statement refers to a specific cross-linking or deposition service, that service should be identified; otherwise the claim should be removed or qualified. This matters because the abstract advertises the addition of ICSD and CCDC numbers as a headline feature.","section":"Section I.A"},{"comment":"No validation protocol for the 344 new entries is described: there is no statement of how each structure was matched to its ICSD/CSD source, how the AFLOW label and Wyckoff assignments were verified, or what automated consistency checks were run. For a database paper whose value is curatorial reliability, this omission is consequential; the P42mc/205 mismatch above is an instance of the kind of error such checks would catch. I request a validation section describing the pipeline and reporting pass/fail rates, or a statement that every new entry was checked by a symmetry-finding tool and by an independent space-group assignment.","section":"Section I.A and Appendix A"}],"minor_comments":[{"comment":"The caption reads 'FIG. 2.aflow.orgtutorial page available in the prototype encyclopedia' and is missing a space; it should read 'aflow.org tutorial page'.","section":"Figure 2 caption"},{"comment":"The phrase 'space groupP2(#3)' is missing a space and should read 'space group P2(#3)'.","section":"Section I.A"},{"comment":"The Jmol command examples contain inserted spaces, such as 'af low li b' and 'af lo wli b', which would prevent copy-paste execution; these should be presented as monospace code blocks without breaks.","section":"Section II (Jmol examples)"}],"recommendation":"major_revision","confidential_remarks":null},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Quick take: this is a straightforward data-descriptor update to a heavily used crystallographic encyclopedia. The 344 new prototypes and the software survey are genuinely useful. But the appendix has at least one internally inconsistent label, which tells me the new-entry list has not been machine-checked, and the authors make an unsupported claim about ICSD/CCDC deposition numbers. Neither flaw sinks the paper, but both need fixing before I'd treat the new entries as fully reliable.\n\nWhat's actually new: 344 prototype entries, bringing the library to 2,127, with a deliberate effort to fill sparse space groups like P2. The ICSD/CCDC cross-references and the new tutorials are real additions. The software survey covering OpenKIM, NOMAD, pymatgen, robocrystallographer, Jmol, and the ML generators is a convenient and, as far as I can tell, accurate map of the ecosystem. The data are public on GitHub under Apache 2.0, so the resource itself is reproducible.\n\nSoft spots, in proportion. The one that matters: in Appendix A, under the header 'P42mc(105)', the new entry 'Br2H12MgO12' is labeled 'A2B12CD12_cP108_205_c_2d_a_2d-001'. The '205' in the label is space group Pa-3, a cubic group, while the section header is P42mc, tetragonal. Different Bravais lattice entirely. This is not a typo in a Wyckoff letter; the label and the header cannot both be right. It could be a typesetting error in the paper rather than in the online database, but either way the published list is not internally consistent. That matters because the AFLOW label is exactly the artifact downstream tools like XtalFinder and pymatgen use. The paper also provides no validation of the 344 new entries—no symmetry check, no comparison against source ICSD/CSD entries. One bad entry doesn't mean many are bad, but it demonstrates that QA is absent.\n\nSecond: the statement that all ICSD entries have a corresponding entry in the CCDC is stronger than what I know to be true, and it is cited to the CCDC homepage rather than to documentation that supports it. Should be corrected or qualified.\n\nMinor: the software survey gets a little list-like and leans on AFLOW's own prior papers, but that is normal for this kind of ecosystem review.\n\nWho it's for: computational materials scientists and informatics tool builders who use the AFLOW prototype library. The paper is a legitimate extension of an established resource, not a conceptual advance. I would send it to review: a referee can quickly sample new entries for label/header consistency and ask for a machine-readable validation appendix. I would not desk-reject it, but I would not accept it as-is.","headline":"Useful data-library update with a real QA gap: one appendix entry contradicts its own space-group header, so the new entries need validation before I'd trust them.","tokens_in":71003,"tokens_out":2067,"would_cite":true,"duration_ms":22702,"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 AFLOW Library of Crystallographic Prototypes now contains 2,127 curated crystal structures, 344 more than the previous edition.","keywords":["crystallographic prototypes","AFLOW prototype label","materials database","space groups","Wyckoff positions","crystallography tutorials","high-throughput materials science","structure identification"],"falsifier":"Take the deposited CIF files for a random sample of the 344 new prototypes, run an independent symmetry finder, and check whether the space group and Wyckoff positions in the AFLOW label match; any mismatch would falsify the claim that the entries faithfully encode the experimental structures.","tokens_in":1379,"feed_emoji":"💎","tokens_out":1551,"duration_ms":82305,"temperature":0.7,"pith_summary":"This paper reports the fifth update to the AFLOW encyclopedia of crystallographic prototypes. It claims that 344 new prototype entries were added, bringing the total to 2,127, and that each new or updated entry now carries a cross-reference number from the ICSD or CCDC databases. The additions were chosen partly to fill gaps in poorly represented space groups, such as P2 (#3), and partly for pedagogy, with many silica and aluminum phosphate structures. The paper also argues that the AFLOW prototype label has become a working interchange format across materials science software, and it documents that use in databases, visualization tools, and machine-learning generators. If right, the library is a standard reference point for connecting experimental crystal structures to computational workflows.","feed_headline":"344 new prototypes push crystal encyclopedia to 2,127","feed_subtitle":"Standard labels now tie experimental deposition numbers to high-throughput materials calculations.","key_machinery":"The central object is the AFLOW prototype label, a machine-readable string such as AB_cF8_225_a_b that records the stoichiometry type, Pearson symbol, space group number, and occupied Wyckoff sites of a structure. The label does the work: it turns a crystal structure into a stable, comparable identifier that can be stored in a database, looked up by software, and cross-referenced against experimental deposition numbers. The encyclopedia's new ICSD and CCDC numbers attach to this label, and the software survey in Section II is organized around programs that accept, search, or generate such labels.","core_discovery":"The central claim is that the AFLOW prototype encyclopedia can be treated as a comprehensive, standardized catalog of crystal structures: it now indexes 2,127 prototypes, and every entry is labeled by a compact string that encodes the arrangement of atoms. The 344 new entries extend coverage into space groups that were previously sparse, with the monoclinic P2 (#3) group given as an example. The site now records ICSD and CCDC deposition numbers that link each prototype to its experimental source. The paper further claims that the prototype label is not just an internal cataloging device but an interoperability standard: it is used as a search field in a major federated materials database, mapped through a standard API, matched by materials-analysis libraries, and consumed by structure generators. Accompanying the data are video and written tutorials on crystallography for computational materials scientists.","pith_inferences":["A natural next step the paper does not take is to measure what fraction of all experimental entries are represented by the 2,127 prototypes; that coverage statistic would turn the catalog from a curated list into a quantitative claim about completeness.","Because the label is symmetry-exact, machine-learning generators could use the new entries to sample isopointal structures systematically, potentially finding stable chemistries that do not appear in the training set.","The inclusion of CCDC numbers for organic and metal-organic entries hints that a parallel taxonomy for molecular crystals could be built on the same label system.","The lack of an automated validation step means the entire update rests on manual curation; an independent re-derivation of space groups from the source depositions would settle whether the 344 additions are reproducible."],"forward_implications":["If the new entries are accurate, the 2,127-prototype library can serve as a near-complete reference for matching experimental crystals to standardized labels.","The P2 gap-filling examples indicate that coverage of the 230 space groups is becoming more even, which supports statistics on how real structures are distributed across space groups.","Cross-linked ICSD and CCDC numbers make each prototype traceable to its original experimental source.","The new tutorials give computational materials scientists a direct path from crystallography basics to using prototype labels.","The documented software ecosystem means a label assigned in one tool can be read by another, easing data exchange."],"supporting_citations":[{"why":"Defines the prototype label and encyclopedia format that this installment extends.","marker":"[6]"},{"why":"The previous installment, giving the entry count and structure that the 344 additions build on.","marker":"[9]"},{"why":"Supplies the experimental deposition numbers now attached to prototype entries.","marker":"[13]"},{"why":"Supplies the CCDC deposition numbers for organic and metal-organic entries.","marker":"[20]"},{"why":"Documents an interatomic-model knowledgebase that categorizes materials by prototype label.","marker":"[26]"},{"why":"A web database that accepts prototype labels as a search field.","marker":"[28]"},{"why":"The API standard used to map prototype labels across materials databases.","marker":"[29]"},{"why":"A materials-analysis library whose matcher returns the prototype label for a structure.","marker":"[30]"},{"why":"The tool that first paired an internal symmetry routine with prototype identification.","marker":"[47]"}],"fun_headline_variants":["2,127 crystal prototypes: AFLOW adds 344 new entries","AFLOW prototype library grows to 2,127 structures","Crystal encyclopedia: 344 new prototypes, 2,127 total","AFLOW's crystal catalog reaches 2,127 with new additions","AFLOW adds 344 prototypes, links to experimental data"],"cache_read_input_tokens":73088,"weakest_assumption_plain":"Everything rests on the 344 new entries being faithful transcripts of their experimental sources; the paper lists the entries but gives no independent check that the assigned space groups and atomic positions are correct.","fun_headline_variants_meta":{"raw":{"variants":["2,127 crystal prototypes: AFLOW adds 344 new entries","AFLOW prototype library grows to 2,127 structures","Crystal encyclopedia: 344 new prototypes, 2,127 total","AFLOW's crystal catalog reaches 2,127 with new additions","AFLOW adds 344 prototypes, links to experimental data"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000289,"raw_usage":{"total_tokens":1600,"prompt_tokens":760,"completion_tokens":840,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":376,"completion_tokens_details":{"reasoning_tokens":749}},"tokens_in":376,"tokens_out":840,"duration_ms":7974,"temperature":1.0,"reasoning_tokens":749,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T15:19:40.852483+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take the deposited CIF files for a random sample of the 344 new prototypes, run an independent symmetry finder, and check whether the space group and Wyckoff positions in the AFLOW label match; any mismatch would falsify the claim that the entries faithfully encode the experimental structures.","supporting_citations":[],"review_version":2}