SQLite terminology databases used by Convert-Pheno.

The databases are organized as a versioned bundle:

```
share/db/
  README
  manifest.json
  concepts_candidates_2_exposure.csv
  v0/
    cdisc.db
    hpo.db
    icd10.db
    ncit.db
    ohdsi.db
    omim.db
```

`manifest.json` records the bundle contract, source snapshots, row counts, and
checksums. Its `currentBundle` field selects the directory used at runtime. A
manifest pointer is used instead of a `current` symbolic link so installed CPAN
distributions behave consistently on Unix and Windows.

The smaller databases are bundled with the distribution. `ohdsi.db` is
distributed separately because of its size. Download it into the selected
bundle directory, currently `share/db/v0/`, or keep it elsewhere and pass the
containing directory with `--path-to-ohdsi-db`. Do not place it directly in
`share/db/`, because runtime database lookup follows `currentBundle`.

Download sources of raw data.

# NCIT
https://bioportal.bioontology.org/ontologies/NCIT

Version Released        Uploaded        Downloads
22.06d (Parsed, Indexed, Metrics, Annotator)    06/30/2022      06/30/2022      OWL | CSV | RDF/XML | Diff

# ICD-10
Downloaded Jul-30-2025 
https://ftp.cdc.gov/pub/Health_Statistics/NCHS/Publications/ICD10CM/2026/icd10cm-table%20and%20index-2026.zip
Version date:  6/12/2025
unzip *zip
icd10cm-tabular-2026.xml (created 

# CDISC Terminology
Downloaded Jan-24-2023
- https://www.cdisc.org/standards/terminology/controlled-terminology
wget https://evs.nci.nih.gov/ftp1/CDISC/SDTM/SDTM%20Terminology.txt
dos2unix SDTM%20Terminology.txt

# OMIM
Downloaded May-22-2023
wget https://raw.githubusercontent.com/macarthur-lab/omim/master/data/full_omim_table.txt
dos2unix full_omim_table.txt

# HPO (Prefix eq 'HP:') 
Downloaded Apr-14-2023
https://hpo.jax.org/app/data/ontology
- v2023-04-05

# OHDSI Standardized Vocabularies
Downloaded from https://athena.ohdsi.org/
Athena vocabulary release: v5.0 29-AUG-22

The release was obtained from the vocabulary_version entry in VOCABULARY.csv.
The current database is approximately 3.2 GB. In addition to concept labels and
identifiers, it contains OMOP domain and standard-concept metadata plus active
`Maps to` and `Maps to value` relationships. These fields are required for
domain-aware standard concept resolution when OMOP is the output model.

Convert-Pheno rejects older four-column copies because they cannot safely
resolve standard concepts for OMOP output.
