vbo-ontology

# Vertebrate Breed Ontology (VBO) ## Overview The Vertebrate Breed Ontology (VBO) is a standardized, structured taxonomy of animal breeds and varieties across 49 domesticated and semi-domesticated species. Developed and maintained by the VBO consortium (veterinary geneticists, bioinformaticians, breed club representatives), VBO enables consistent breed naming, automated breed-predisposition reasoning, and integration with genetic disease databases (OMIA, VetClaw). With 19,500+ breed concepts and hierarchical classification, VBO powers precision veterinary medicine. ## When to Use - User identifies patient breed and needs standardized breed nomenclature for records/research - User queries breed-specific health predispositions (MDR1 in herding breeds) - User resolves breed name ambiguities (e.g., "English Springer Spaniel" vs. "Springer Spaniel" variant) - User performs cross-species breed classification (dog breed groups, horse registry standards) - User links breed-specific genetic test recommendations to patient phenotype - Keywords: VBO, breed ontology, breed identification, breed predisposition, standardized nomenclature, FCI ## What is VBO? **Definition:** Formal ontology structured as directed acyclic graph (DAG) where: - Nodes = breed concepts (unique IDs) - Edges = hierarchical relationships (is-a, part-of) - Metadata = breed names (preferred + synonyms), species, FCI classification (for dogs/cats), breed clubs **Coverage (by Species):** - **Companion Animals:** Canis familiaris (dog) - 900+ breeds; Felis catus (cat) - 250+ breeds; Rabbit (200+), Guinea pig (50+) - **Livestock:** Cattle (300+), Horse (350+), Sheep (500+), Goat (200+), Swine (100+), Poultry (200+) - **Exotic Species:** Fish (200+), Rodents, Amphibians, Reptiles **Total:** 19,500+ breed concepts across 49 species ## VBO Hierarchical Structure **Dog (Canis familiaris) Example:** ``` VBO_DOG_ROOT (Dog - all breeds) ├── VBO_GROUP_1 (FCI Group 1: Sheepdogs & Cattle Dogs) │ ├── VBO_COLLIE (Rough Collie) │ ├── VBO_COLLIE_SMOOTH (Smooth Collie) │ ├── VBO_AUSTRALIAN_SHEPHERD (Australian Shepherd) │ ├── VBO_GERMAN_SHEPHERD (German Shepherd Dog) │ └── ... (50+ herding breeds) ├── VBO_GROUP_2 (FCI Group 2: Pinschers, Schnauzers, Molossians) │ ├── VBO_BOXER (Boxer) │ ├── VBO_DOBERMAN (Doberman Pinscher) │ └── ... (20+ molossoid breeds) ├── VBO_GROUP_3 (FCI Group 3: Terriers) │ ├── VBO_SMOOTH_FOX_TERRIER (Smooth Fox Terrier) │ └── ... (30+ terrier breeds) ├── VBO_GROUP_8 (FCI Group 8: Retrievers, Flushing Dogs) │ ├── VBO_LABRADOR (Labrador Retriever) │ ├── VBO_GOLDEN (Golden Retriever) │ └── ... (30+ gun dogs) └── ... (8 total FCI groups + miscellaneous/non-FCI breeds) ``` **Cattle (Bos taurus) Example:** ``` VBO_CATTLE_ROOT ├── VBO_DAIRY (Dairy Breeds) │ ├── VBO_HOLSTEIN (Holstein-Friesian) │ ├── VBO_JERSEY (Jersey) │ ├── VBO_GUERNSEY (Guernsey) │ └── VBO_BROWN_SWISS (Brown Swiss) ├── VBO_BEEF (Beef Breeds) │ ├── VBO_ANGUS (Angus) │ ├── VBO_HEREFORD (Hereford) │ ├── VBO_CHAROLAIS (Charolais) │ └── ... └── VBO_DUAL_PURPOSE (Dairy/Beef Breeds) ├── VBO_SIMMENTAL (Simmental) └── ... ``` ## VBO Breed ID Format **Standard Format:** `VBO_XXXXXX` (6-character code or full name reference) **Examples:** - VBO_0000147 = Labrador Retriever (canine) - VBO_0000456 = Collie (canine) - VBO_0000123 = Siamese (feline) - VBO_0001000 = Holstein (bovine) - VBO_0002156 = Thoroughbred (equine) **Nomenclature Standards:** - Preferred name = FCI standard (for FCI-recognized breeds) - Synonyms = regional variants, historical names, abbreviated forms - Example: "Rough Collie" (preferred) ≈ "Lassie type" (historical) ≈ "Collie long-coat" (variant) ## Accessing VBO **Web Interface (OBO Foundry/BioPortal):** https://www.ebi.ac.uk/ols/ontologies/vbo (EMBL-EBI OLS - Ontology Lookup Service) **Search Functions:** 1. **Breed by Name:** Enter "Labrador" → Returns VBO_0000147 + linked data 2. **Breed by FCI Code:** Enter "8.1" (FCI retriever code) → Lists all retrievers 3. **By Species:** Select "Canis familiaris" → Browse all dog breeds 4. **By Predisposition:** (Advanced) Query "MDR1 mutation" → Returns all affected breeds **Download Formats:** - OWL (Web Ontology Language): Full semantic structure, machine-readable - OBO Format: Text-based, human-readable - JSON: Programmatic access - CSV: Breed list spreadsheets **File Location:** https://github.com/vbreeds/vbo (open-source repository) ## VBO-OMIA Integration **Power of Linking:** VBO enables automated health predisposition reasoning by linking breed ID to inherited disease databases (OMIA). **Example Query:** ``` Breed: VBO_0000456 (Collie) → Link to OMIA records for Collie predispositions → Results: - MDR1 mutation: 50-70% carriers - Collie Eye Anomaly: 15-25% carriers - Progressive Retinal Atrophy: Breed-specific loci - Microphthalmia: Rare → Veterinarian alert: "Collie breed: avoid ivermectin (MDR1); screen eyes (CERF)" ``` **Breeding Program Integration:** ``` 1. Breeder registers breeding dogs: VBO ID + genetic test results (OMIA-linked) 2. System identifies mating risks: - Both parents carriers of PRA → 25% affected puppies - One parent MDR1 homozygous, other carrier → all puppies carriers 3. Recommendation: Avoid mating; counsel on genetic implications ``` ## VBO Hierarchical Relationships **Parent-Child Relationships:** ``` FCI Group 8 (Retrievers) ├── Labrador Retriever (VBO_0000147) │ ├── Black Labrador (phenotypic variant) │ ├── Yellow Labrador (phenotypic variant) │ └── Chocolate Labrador (phenotypic variant) └── Golden Retriever (VBO_0000234) ``` **Breed Variants (Size/Coat):** Some breeds have recognized size/coat variants (not separate breed in FCI, but distinct lines): - **Standard Poodle** vs. **Miniature Poodle** vs. **Toy Poodle** - VBO distinguishes size variants with separate IDs - Different health predispositions (patella luxation more common in toy sizes) - **Rough Collie** vs. **Smooth Collie** - Both VBO IDs, same genetic predispositions (MDR1) - Phenotypic difference (coat length) only ## Breed Predisposition Reasoning **Automated Workflow (VetClaw AI):** ``` 1. Patient Information Input: - Species: Canis familiaris - Breed: "Rough Collie" - Age: 8 months - Gender: Female 2. VBO Lookup: - Resolves "Rough Collie" → VBO_0000456 - Identifies parent FCI group: Group 1 (Herding Dogs) 3. OMIA Predisposition Query: - Search OMIA: "VBO_0000456" + inherited diseases - Returns: MDR1, Collie Eye Anomaly, PRA, etc. 4. Genetic Test Recommendations: - MDR1: Recommended (drug interaction implications) - Eye exam: CERF certification (Collie Eye Anomaly screening) - PRA testing: Candidate gene panel 5. Clinical Alert Generation: - "Breed predisposition: Avoid ivermectin + analogs" - "Recommend ophthalmology exam before breeding" 6. Treatment Plan Modification: - Alternative antiparasitic: Spinosad (not MDR1-substrate) - Genetic counseling: If breeding planned ``` ## Cross-Species Breed Standardization **Problem:** Same breed name across species with different meanings - "Angora": Could be Angora rabbit, Angora goat, or Angora cat - VBO resolves with species prefix: VBO_ANGORA_RABBIT vs. VBO_ANGORA_GOAT **International Breed Name Variation:** - "English Springer Spaniel" (FCI) vs. "English Springer Spaniel" (AKC) vs. "Springer Spaniel" (colloquial) - VBO preferred: FCI standard; synonyms capture variants **Horse Registry Integration:** - Thoroughbred (Equestrian registry: Jockey Club) - Quarter Horse (AQHA registration) - VBO links both to equine taxonomy, but maintains registry distinctions ## VBO Querying Examples **Example 1: Find All Large-Breed Dogs (Hip Dysplasia Risk)** ``` SPARQL Query: SELECT ?breed ?breedLabel WHERE { ?breed rdfs:subClassOf vbo:GROUP_8 . # Retrievers, gun dogs ?breed rdfs:label ?breedLabel . } Returns: Labrador, Golden Retriever, etc. (all large breeds) Filter by predisposition: High hip dysplasia prevalence ``` **Example 2: Identify All Cat Breeds with PKD1 Risk** ``` Query: "Felis catus" + "PKD1" + inheritance:autosomal_dominant Returns: Persian, Maine Coon, Bengal, British Shorthair, etc. Link to OMIA: PKD1 prevalence per breed Breeding recommendation: Screen parents (ultrasound kidney assessment) ``` **Example 3: Breed Equivalence Across Registries** ``` Query: VBO_LABRADOR (FCI) + AKC equivalent Returns: Labrador Retriever (AKC) = Labrador (FCI) = same breed Use for breed-predisposition research (align datasets from different registries) ``` ## VBO-Based Breed Identification Workflow **Ambiguous Case: User Says "Spaniel"** ``` 1. VBO Query: "Spaniel" (search for partial name match) 2. Returns multiple breeds: - English Springer Spaniel (VBO_0000XXX) - Cocker Spaniel (VBO_0000YYY) - Cavalier King Charles Spaniel (VBO_0000ZZZ) - American Spaniel Variety (multiple lines) 3. Veterinarian Clarifies: - "Which spaniel breed exactly?" - Obtain FCI classification or show pedigree 4. Once Breed Confirmed: - VBO ID assigned (e.g., VBO_0000XXX = English Springer Spaniel) - Health predispositions retrieved - Genetic testing recommendations automated 5. Medical Record: - Store VBO ID (standardized across clinics/research) - Enable data sharing, breeding program tracking ``` ## Limitations - **Non-FCI Breeds:** Many regional/emerging breeds lack VBO classification - **Mixed-Breed Animals:** VBO applies to purebreds; hybrids/mongrels not classified - **Breed Validity:** FCI recognition varies; some "breeds" are phenotypic variants, not genetically distinct - **Phenotypic vs. Genetic:** VBO emphasizes FCI breed standards (phenotype) rather than genetic similarity (genotype) - Example: English Springer Spaniel vs. Cocker Spaniel are distinct FCI breeds but genetically closely related - **Updating Lag:** New breeds recognized by breed clubs; VBO updates lag (6-12 months typical) - **Species Bias:** Dog, cat, horse breeds well-represented; exotic species sparse ## Sources - **VBO GitHub:** https://github.com/vbreeds/vbo - **OBO Foundry:** https://www.obofoundry.org/ontology/vbo - **EMBL-EBI OLS:** https://www.ebi.ac.uk/ols/ontologies/vbo - **FCI (Fédération Cynologique Internationale):** https://www.fci.be (dog breed standards) - **TICA (The International Cat Association):** https://www.tica.org (cat breed standards) - **Equine Breed Registries:** Various (Jockey Club, AQHA, etc.) - **OMIA (Cross-Reference):** https://www.omia.org ## Advanced Integration VetClaw SDK provides VBO breed lookups and OMIA predisposition queries integrated into clinical decision support. See `/sessions/charming-practical-mayer/mnt/OpenVet/vetclaw/vetclaw/sdk/` for Python/TypeScript breed ontology client examples.