Phenofhy

Data model ​

Phenofhy uses a simple naming convention for fields:

  • Raw fields are entity.field (for example, participant.birth_year).
  • Derived fields live under derived.* (for example, derived.age_at_registration).

Metadata dictionaries ​

Phenofhy uses metadata dictionaries to describe fields, coding domains, and entities in the Our Future Health TRE.

For analyses running in the TRE, metadata exported from DNAnexus can include:

  • *.codings.csv for code-to-label mappings
  • *.data_dictionary.csv for field descriptions and metadata
  • *.entity_dictionary.csv for entity metadata

These files are retrieved and processed by pipeline.metadata() for use with project data.

Metadata used for simulation ​

Phenofhy also includes the metadata required by phenofhy.simulate within the installed package. Users do not need to download these files separately.

For example:

from phenofhy import simulate

df = simulate.simulate_phenotype_df(
    sample=500,
    seed=42,
)

The simulation utilities use the packaged data dictionary and coding information to generate synthetic OFH-like phenotype data.

Coding names ​

Coding names link fields in the data dictionary to their permitted coded values.

For example:

  • participant.demog_sex_1_1 maps to coding name DEMOG_SEX_1_1.
  • The corresponding coding domain defines the permitted values for that field.

Phenofhy handles this mapping internally when generating simulated phenotype data.