liveBook · Manning

Phenotypic feature

As explained in Knowledge Graphs and LLMs in Action

A phenotypic abnormality is an observable physical or biochemical characteristic that deviates from typical traits.

A phenotypic feature is an observable physical or biochemical characteristic. An individual's phenotype is the sum of the features manifested with a disease; features can also be modeled as diseases, depending on context. section 3.0 section 3.1.1

Use phenotypic features as structured, traceable links between clinical observations and diseases:

  • Represent each standardized feature as an ontology concept with an identifier, label, synonyms, description, and hierarchy. section 3.1.2
  • Connect diseases to features through annotated relationships that preserve evidence, source references, onset, frequency, sex, modifiers, and curation metadata. section 3.1.2 section 3.3.2
  • Query direct matches or traverse subclass relationships to retrieve diseases associated with specific or broader phenotype categories. section 3.4 section 3.5

Type 1 diabetes as a context-dependent feature

Type 1 diabetes mellitus uses HP:0100651 when represented as a phenotypic feature and OMIM:222100 when represented as a disease. Querying additional features such as growth delay, large knee, sensorineural hearing impairment, and pruritus can surface candidate rare diseases, including Ondontochondrodysplasia 2 with hearing loss and diabetes. section 3.1.1 section 3.4

Figure 3.4 Type 1 diabetes mellitus can be considered either a disease or a phenotypic feature. Based on the context, two different IDs can be adopted.
figure

Shows why the same clinical concept can require different identifiers when modeled as a disease or as a phenotypic feature. section 3.1.1

How Knowledge Graphs and LLMs in Action covers it

How deep it goes
Defines features, models annotated relationships, and demonstrates direct and inferred queries; it stops short of clinical validation or treatment decisions.
Examples
Uses Type 1 diabetes, endocrine-system abnormalities, growth delay, hearing impairment, pruritus, and HPO disease associations.
What you can do after
Diagnose with phenotype-based Cypher queries and ontology inference that rank diseases by matching observed features.
How it connects
Connects phenotype, disease, ontology, knowledge graph, RDF, OWL, provenance, and semantic inference.

Related concepts

This explainer was generated by Manning's knowledge graph from the text of Knowledge Graphs and LLMs in Action. The section links point to the parts of the book it draws on.