{"id":"b3695a7a-78e9-4b97-af09-a9e1ff6f0ae6","arxiv_id":"2501.01454","paper_version":3,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"The paper describes a fourfold suite of pathogen reference ontologies (VIDO, BIDO, MIDO, PIDO) extending the Infectious Disease Ontology with modular, reusable semantics.","lead":"Researchers created four pathogen-specific extensions of the Infectious Disease Ontology (VIDO for viruses, BIDO for bacteria, MIDO for fungi, PIDO for parasites), all built around a shared hub-and-spoke design. The suite standardizes how infectious disease data is organized and could make it easier to integrate data across different pathogen types.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"MIDO and PIDO define 'infectious disease course' as a disease rather than as a process, breaking the IDO pattern that the suite claims to inherit.","rationale":"The reader's weakest assumption concerns whether the IDO/BFO design pattern can represent all four pathogen types without conceptual distortion. My sharper concern is a concrete violation of that pattern in the published definitions: MIDO and PIDO misclassify disease course as disease, while VIDO and BIDO use the correct process-based definition. This is an internal inconsistency in the suite's central design template, not a disagreement with an outside consensus, so it directly threatens the 'rigorous' and 'uniform' claims of the central thesis. It also has a clear, inexpensive test: inspect the OWL files and reason over them, which the reader did not do. The concern does not warrant rejection, because the artifacts are available and the error is localized and likely fixable. It does reinforce the CONDITIONAL verdict: correction of these definitions, and ideally a demonstration that the corrected axioms are logically coherent, should precede claims of modularity and reuse. I mark agreement as partial because the reader noted the disease-course inconsistency in the rationale but framed the weakest assumption more broadly as a question of template adequacy for toxin-mediated and parasite-stage phenomena.","tokens_in":37871,"tokens_out":3535,"duration_ms":40145,"concrete_test":"Download the released MIDO and PIDO OWL artifacts from the GitHub repositories and load them into Protégé with HermiT or Pellet. Inspect the asserted axioms for 'fungal infectious disease course' and 'parasite infectious disease course': if either class is declared equivalent to or a subclass of an infectious disease, or if its only formal definition is a textual annotation stating that it is a disease, the error is confirmed. Then compare these axioms with the VIDO and BIDO definitions to verify the discrepancy, and run a reasoner query to see whether any disease-course individual would be inferred to be a disease.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The suite's central value proposition is a rigorous, uniform intermediate layer between IDO and pathogen-specific ontologies. The Methods section explicitly adopts the IDO/OGMS disease-course pattern, where an infectious disease course is a process that realizes an infectious disease. Tables 2 and 3 (VIDO, BIDO) follow this pattern. Tables 4 and 6 do not: MIDO defines 'fungal infectious disease course' as 'Infectious disease whose physical basis is a fungal disorder...' and PIDO defines 'parasite infectious disease course' as 'Infectious disease whose physical basis is a parasite disorder...'. This conflates a disease (a disposition) with a disease course (a process), a category mistake that materially weakens the claim that the suite provides rigorous ontological representations. Because modularization and reuse depend on downstream ontologies importing a consistent, well-founded reference layer, the error in half of the spokes is not cosmetic: it would propagate an inconsistent disease-course semantics into any domain ontology built on MIDO or PIDO. The paper itself acknowledges that implementation is future work, but the claimed conceptual foundation is already internally inconsistent.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper introduces a suite of four OBO-style reference ontologies extending the Infectious Disease Ontology (IDO): VIDO, BIDO, MIDO, and PIDO. The stated aim is to provide an intermediate semantic layer between IDO and pathogen-specific domain ontologies, using a hub-and-spoke methodology, with BFO/IDO design patterns and reuse of terms from GO, PRO, ChEBI, MONDO, and others. The authors report ontology size and reuse statistics, describe major classes and design patterns (e.g., virus replication cycle, bacterial toxin vs. infectious disorder, parasite life stages), mention automated reasoning checks with HermiT and Pellet, and outline CI/CD-based governance and future extension plans. The central claim is that these reference ontologies advance modularization and reusability within the IDO ecosystem.","tokens_in":38029,"tokens_out":1935,"duration_ms":22620,"significance":"If the central claim is substantiated, the suite would be a useful contribution to biomedical ontology infrastructure: it responds to a real need for consistent mid-level pathogen-specific classes, the artifacts are openly available, the reasoning checks with HermiT and Pellet provide some machine-checked assurance, and the proposed CI/CD pipeline and GitHub governance are good practical steps. The paper also gives appropriate credit to existing resources (IDO, OGMS, MONDO, OPL, etc.). However, the paper currently provides no quantitative or comparative evidence that the suite improves modularity, reuse, or data integration; the claim is asserted rather than measured. In addition, a category-level definitional inconsistency in two of the four ontologies undermines the suite's internal coherence and its claim to provide a rigorous uniform reference layer.","major_comments":[{"comment":"The definitions of 'fungal infectious disease course' (Table 4) and 'parasite infectious disease course' (Table 6) classify these terms as diseases: e.g., 'Infectious disease whose physical basis is a fungal disorder...' and 'Infectious disease whose physical basis is a parasite disorder...'. This contradicts the OGMS/IDO pattern explicitly adopted in Methods and correctly followed in Tables 2 and 3, where an infectious disease course is an infectious disease course that is the realization of an infectious disease (a process, not a disposition). The error conflates a disease (disposition) with a disease course (process) and would propagate a semantically inconsistent disease-course hierarchy to any downstream ontology importing MIDO or PIDO terms. The definitions should be corrected to the pattern used in VIDO and BIDO before the suite can be presented as a coherent reference layer.","section":"Tables 4 and 6; Results (MIDO, PIDO)"},{"comment":"The central claim that the suite 'advances modularization and reusability' is asserted rather than demonstrated. The paper provides no modularity metrics, no comparison with alternative modeling strategies, no analysis of imported-term redundancy or overlap, and no downstream application that measures reuse. The Discussion explicitly defers empirical investigation to future work ('subsequent empirical investigation'), and the Conclusion frames the value as a 'conceit'. This makes the headline conclusion currently unsupported. To substantiate the claim, the authors should either report concrete reuse or interoperability evidence (e.g., refactoring of IDOFLU/IDOMAL/IDOSCHISTO using the new reference layer, or quantitative term-reuse statistics across the ecosystem) or explicitly soften the conclusion to a design proposal.","section":"Discussion and Conclusion; Abstract"},{"comment":"The BIDO distinction between 'bacterial toxin disorder' and 'bacterial infectious disorder' is well motivated, but the presentation leaves ambiguous whether 'bacterial infectious disease course' and 'bacterial pathogenesis' are asserted to be disjoint or partially overlapping. Table 3 defines 'bacterial infectious disease course' as the realization of a bacterial infectious disease, while 'bacterial pathogenesis' is defined without requiring an infection process. Since a bacterial toxin-induced non-infectious disease can have a disease course but is not an infectious disease course, the paper should make explicit whether a 'bacterial infectious disease course' can realize a disease whose basis is a bacterial toxin disorder that is not an infectious disorder. The current text suggests these are cleanly separated but does not state the logical relationship, which matters for downstream reasoning.","section":"Methods; Results (BIDO)"}],"minor_comments":[{"comment":"Several typographical and grammatical errors should be corrected: 'an reference ontology' (Results, VIDO section), 'subject-matter matters' (Methods), 'A Fourfold Pathogen Reference Ontology Suite' repeated in the running title, and inconsistent capitalization of 'Fungus' and 'Parasite' in class definitions.","section":"Throughout"},{"comment":"References 50–52 duplicate references 12, 13, and 7 with different numbering; the reference list should be deduplicated and all in-text citations rechecked. Also, some references cite 'Retrieved December 24, 2024' and 'Retrieved April 17, 2025' inconsistently; the access dates should be harmonized.","section":"References"},{"comment":"The two SWRL rules in the VIDO section are presented in a way that is hard to read; the authors should consider aligning the rule syntax and explaining in prose that the rules are applied only to virions participating in the same replication cycle. Also, the paper states that OWL 'is not amenable to representing conditional scenarios' but then uses SWRL; this contrast could be clarified, since SWRL rules are themselves not part of OWL 2 DL semantics and may affect decidability.","section":"Methods; SWRL rules"},{"comment":"In the MIDO section, the axiom for 'opportunistic fungal pathogen' is written as an 'inheres in some fungus' axiom, but the class name suggests it classifies fungi, not dispositions. This should be rephrased or checked in the OWL artifact to avoid a misleading presentation of what the axiom asserts.","section":"Results (MIDO)"},{"comment":"The GitHub URLs for the four ontologies are given, but the paper does not specify version identifiers, release dates, or a hash/DOI for the exact artifacts that were reasoned over with HermiT and Pellet. Adding persistent versioning information would improve reproducibility of the reported reasoning checks.","section":"Availability of Data and Materials"}],"recommendation":"major_revision","confidential_remarks":"The paper is a useful infrastructure report but currently sells itself as demonstrating modularization and reusability when it actually proposes them. The definitional disease-course error in MIDO and PIDO is the kind of issue that, if left in the released ontologies, would be amplified by reuse; it is fixable, but it is load-bearing for the 'rigorous' and 'uniform' claims. I would not go so far as reject, because the artifacts are genuinely available, the basic methodology is sound, and the error is localized rather than systemic. The authors should also be encouraged to report at least one concrete reuse or refactoring exercise, even small-scale, to ground the central claim."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"This paper does something real: it adds three new pathogen-specific reference ontologies (BIDO, MIDO, PIDO) to the IDO family, with publicly available OWL artifacts, and it follows the hub-and-spoke methodology with heavy reuse of existing OBO terms. VIDO is largely a summary of prior work, but the three new ontologies contain dozens of defined classes and logical axioms. The authors ran HermiT and Pellet over the artifacts and report the results. That is genuine infrastructure work, honestly presented as laying foundations rather than claiming finished applications.\n\nThe stress-test concern is accurate. In Tables 4 and 6, 'fungal infectious disease course' and 'parasite infectious disease course' are defined as 'Infectious disease whose physical basis is...' That makes the class a disease (a disposition), not a course (a process). This directly contradicts the IDO/OGMS pattern the paper says it inherits, where an infectious disease course is the realization of an infectious disease. VIDO and BIDO get this right; MIDO and PIDO do not. Since the suite's value proposition is a consistent intermediate layer, this is more than cosmetic. Any domain ontology importing MIDO or PIDO would inherit a conflated disease/course semantics for two of the four spokes. It is fixable, but it should be fixed before downstream use.\n\nThe bigger soft spot is the central claim about modularization and reusability. The paper asserts this benefit repeatedly, but nowhere measures or demonstrates it. The Discussion acknowledges that implementation and empirical evaluation are future work, and the Limitations section is upfront about governance and update challenges. Still, the abstract and conclusions state the benefit as fact. A single worked example—say, refactoring IDOMAL or BCIDO to import PIDO or BIDO—would materially strengthen the case.\n\nMinor issues include a handful of citation inconsistencies and typos, but nothing that changes the substance.\n\nWho gets value from this paper? Ontology engineers working in the IDO/OBO ecosystem, and researchers who need a stable middle layer for pathogen-specific data curation. The paper is not a scientific discovery, but it is a serious engineering report with reusable artifacts. I would send it to peer review, with the disease course error flagged as a required revision and the modularity claim asked to be supported with at least one concrete alignment or reuse example. I would bring it to a reading group focused on biomedical ontology practice, though not for a general audience.","headline":"Useful new reference ontologies for the IDO ecosystem, but a definitional slip in MIDO and PIDO disease courses undercuts the suite's core design pattern.","tokens_in":38610,"tokens_out":1862,"would_cite":false,"duration_ms":20105,"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 claims that four pathogen-specific reference ontologies—VIDO, BIDO, MIDO, and PIDO—form a needed intermediate layer between the general Infectious Disease Ontology and disease-level ontologies.","keywords":["infectious disease ontology","pathogen reference ontology","hub-and-spoke methodology","VIDO","BIDO","MIDO","PIDO","ontology modularization"],"falsifier":"An automated reasoning test on BIDO that classifies a stated case of non-infectious food-borne botulism as a bacterial infectious disorder would show the assumed template distorts the domain; similarly, a PIDO alignment of the existing malaria and schistosomiasis ontologies that cannot represent the parasite vector as a host-pathogen-vector tripartite relation would falsify the suite's claim of guarding parasite domain modeling.","tokens_in":37655,"feed_emoji":"🦠","tokens_out":7498,"duration_ms":71354,"temperature":0.7,"pith_summary":"The paper claims that infectious disease data needs an intermediate semantic layer between the very general Infectious Disease Ontology (IDO) and ontologies for particular diseases, and that this layer is best built as four pathogen-specific reference ontologies: one each for viruses, bacteria, fungi, and parasites. Following a 'hub-and-spoke' methodology, each ontology imports shared terms from IDO and adds pathogen-specific classes defined with the pattern 'A is a B that C's.' The four ontologies—VIDO, BIDO, MIDO, and PIDO—are meant to act as guardrails so that disease-level ontologies can be built faster, stay aligned with each other, and reuse vetted definitions rather than starting from scratch. A sympathetic reader would care because this is a concrete proposal for making infectious-disease datasets comparable across pathogen types, with the COVID-19 experience as the motivating case for updating and modularizing the IDO ecosystem.","feed_headline":"Four new ontologies put a shared layer under disease data","feed_subtitle":"Virus, bacteria, fungus, and parasite ontologies make infectious-disease datasets interoperable and reusable.","key_machinery":"The load-bearing mechanism is the 'hub-and-spoke' design pattern combined with the IDO/OGMS definition template 'A is a B that C's': each new class is defined as a subclass of a parent imported from IDO or another ontology, with a differentiating clause specifying what makes it specific. Around this template, the suite builds the disorder-infection-disease-disease-course spine inherited from IDO, and adds pathogen-type classifications based on dispositions (e.g., infectious structure as acellular structure bearing an infectious disposition; pathogen as material entity bearing a pathogenic disposition). OWL 2 with automated reasoning, plus SWRL rules for conditional stage ordering in VIDO and equivalency axioms such as the pilus axiom in BIDO, turn these definitions into inferable classifications rather than merely labeled hierarchies.","core_discovery":"The central claim is that creating pathogen-specific reference ontologies advances modularization and reusability within the IDO ecosystem, with each of VIDO, BIDO, MIDO, and PIDO occupying a 'middle' position: more specific than IDO, more general than a disease-level ontology. The paper reports the scope, major classes, and relations of each ontology, and shows how they are assembled by reusing terms from IDO and other ontologies and adding new defined classes. Concretely, VIDO classifies viruses by Baltimore class and by infectious disposition, and models the virus replication cycle with SWRL rules that order stages conditionally; BIDO distinguishes bacterial toxin disorder from bacterial infectious disorder so that food-borne botulism is not misclassified as an infection; MIDO introduces yeast, mold, and dimorphic fungus, and formalizes opportunistic fungal pathogen with an axiom restricting realization to non-immunocompetent hosts; PIDO separates infection-causing from disease-causing parasite life stages, which lets one model diseases like schistosomiasis where infection and disease are caused by different stages. The paper concludes that these four ontologies provide 'guardrails' that save researchers time and effort when building domain ontologies, just as libraries save programmers from rewriting common code.","pith_inferences":["The BIDO distinction between bacterial toxin disorder and bacterial infectious disorder suggests a reusable template for non-infectious toxin-mediated diseases generally, which could be extended to food safety and environmental health ontologies.","The VIDO SWRL pattern for conditional stage ordering is a general trick for representing branching life cycles in OWL, and could be carried over to parasite vector cycles where the same pathogen has different hosts at different stages.","If the suite is adopted, one measurable payoff is reduced definitional drift: aligned domain ontologies should share more term reuse and fewer conflicting definitions than the legacy extensions, a claim future studies could test.","The coordination infrastructure described (a shared GitHub organization, CI/CD, term-request workflows) implies the suite's real product is not just four ontologies but a governance process for keeping them consistent, which is where the project's long-term value would lie."],"forward_implications":["Legacy IDO extensions such as the influenza, HIV, dengue, staphylococcus, brucellosis, meningitis, malaria, and schistosomiasis ontologies can be refactored to import from the matching reference ontology, bringing them into alignment with updated IDO and BFO.","Researchers building a new pathogen-specific disease ontology can start from vetted intermediate classes instead of modeling from the top down, which the paper argues saves time and reduces errors.","The suite's common definitions for infection, infectious disorder, infectious disease, and disease course make it possible to compare and integrate data across virus, bacterium, fungus, and parasite disease domains.","Ontology developers can use automated reasoners to infer classifications, such as recognizing a pilus as both an adhesion factor and a virulence factor, or recognizing a positive-sense RNA virus's translation stage as preceding transcription in a replication cycle."],"supporting_citations":[{"why":"Defines the Basic Formal Ontology classes (material entity, quality, process, role) that provide the top-level semantics for all four ontologies.","marker":"[1]"},{"why":"Supplies the OBO Foundry design principles the suite follows for syntax, identifiers, openness, modularity, and reuse of relations.","marker":"[2]"},{"why":"Documents the IDO update and the need to realign its virus and pathogen extensions after BFO became an ISO standard.","marker":"[3]"},{"why":"Introduces VIDO and the 'guardrails' idea that the other three ontologies model themselves on.","marker":"[16]"},{"why":"Advocates the hub-and-spoke approach to building IDO extensions, the methodology the paper adopts.","marker":"[19]"},{"why":"Originates the OGMS disorder/disease/disease-course framework that IDO's infectious disease definitions inherit.","marker":"[28]"},{"why":"Provides the Baltimore classification used as VIDO's primary virus-classification scheme.","marker":"[37]"},{"why":"The Ontology for Parasite Lifecycle supplies the parasite life stage and parasite organism vocabulary that PIDO extends.","marker":"[79]"}],"fun_headline_variants":["Four pathogen ontologies standardize disease data","New ontologies unify virus, bacteria, fungus, parasite data","Infectious disease ontologies get four new modular branches","Pathogen ontologies save time like code libraries","IDO expands with virus, bacteria, fungus, parasite suites"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The paper assumes that one definitional template built around IDO's disorder and infection classes fits all four pathogen types without distortion, including bacterial toxin-mediated disease that is not an infection and parasite life stages that change phenotype between infection and disease.","fun_headline_variants_meta":{"raw":{"variants":["Four pathogen ontologies standardize disease data","New ontologies unify virus, bacteria, fungus, parasite data","Infectious disease ontologies get four new modular branches","Pathogen ontologies save time like code libraries","IDO expands with virus, bacteria, fungus, parasite suites"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000126,"raw_usage":{"total_tokens":1136,"prompt_tokens":995,"completion_tokens":141,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":611,"completion_tokens_details":{"reasoning_tokens":62}},"tokens_in":611,"tokens_out":141,"duration_ms":2315,"temperature":1.0,"reasoning_tokens":62,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-10T22:55:05.539773+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"An automated reasoning test on BIDO that classifies a stated case of non-infectious food-borne botulism as a bacterial infectious disorder would show the assumed template distorts the domain; similarly, a PIDO alignment of the existing malaria and schistosomiasis ontologies that cannot represent the parasite vector as a host-pathogen-vector tripartite relation would falsify the suite's claim of guarding parasite domain modeling.","supporting_citations":[{"cited_title":"The Brucellosis Infectious Disease Ontology (IDOBRU) [60] 20 A Fourfold Pathogen Reference Ontology Suite","cited_arxiv_id":null,"evidence_quote":"Defines the Basic Formal Ontology classes (material entity, quality, process, role) that provide the top-level semantics for all four ontologies."},{"cited_title":"pathogen","cited_arxiv_id":null,"evidence_quote":"Documents the IDO update and the need to realign its virus and pathogen extensions after BFO became an ISO standard."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Originates the OGMS disorder/disease/disease-course framework that IDO's infectious disease definitions inherit."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the Baltimore classification used as VIDO's primary virus-classification scheme."}],"review_version":1}