🧬
BioKGv1.0 · OWL-based
Alternative to PrimeKG · OWL Ontology-based

BioKGKnowledge Graph

Heterogeneous biomedical knowledge graph constructed from 8 OWL ontologies using SPARQL triple extraction and cross-ontology alignment.

HGNC · OGG · MONDO · NBO · HENEGEO · NPO · NIGO · AIO

35
Total Nodes
across 8 node types
38
Total Edges
17 relation types
📂
8
OWL Sources
ontology files parsed
24
Attributes
unique node properties

3. Node Types (8)

🧬
Gene8 nodes
🦠
Disease6 nodes
💊
Drug5 nodes
🧠
Behavior6 nodes
Neuron4 nodes
🫁
Anatomy2 nodes
🔗
Pathway2 nodes
🔬
Phenotype2 nodes

5. Relationships (17)

associated_with6
driver_mutation2
expressed_in1
causes1
targets2
modulates2
inhibits2
treats5
participates_in2
activates2
regulates2
mediated_by2
regulated_by1
has_phenotype2
is_site_of2
dysregulated_in3
interacts_with1

1. Alternative Development Method

📂
STEP 01
OWL Parsing

Parse RDF/OWL XML with ElementTree or rdflib. Extract owl:Class, rdfs:label, owl:ObjectProperty, rdfs:subClassOf triples.

STEP 02
Triple Extraction

Execute SPARQL SELECT queries to pull (Subject, Predicate, Object) triples. Normalize IRIs across ontologies via shared cross-references.

🔗
STEP 03
Cross-Ontology Align

Match HGNC↔OGG via HGNC IDs. Align MONDO↔HENEGEO via disease xrefs. Link NPO↔NBO via behavior-neuron co-occurrence.

🕸
STEP 04
Graph Build

Construct heterogeneous typed graph. Store as JSON nodes+edges. Validate completeness, connectivity, and consistency.

2. Data Repositories (8 OWL Files)

HGNChgnc.owl
Standardized human gene symbols and names
8
nodes
OGGogg-merged.owl
Biological ontology for genes and genomes
8
nodes
MONDOmondo-simple.owl
Unified cross-species disease ontology
6
nodes
NBOnbo.owl
Human and animal behavioral ontology
6
nodes
HENEGEOHENEGEO_v1_0.owl
Cancer treatment and clinical ontology
5
nodes
NPOnpo.ttl
Neuron type classification ontology
6
nodes
NIGONIGO_vv1.owl
Gene-imaging-neurological links
2
nodes
AIOaio-full.owl
Deep learning networks and AI models
0
nodes