Overview

The purpose of this code is to generate phenotypes in the COHORT table of the UKBB OMOP CDM Postgresql database. Each phenotype will be given a unique cohort definition id, and will be recorded in data/AllCohorts.csv such that there is no overlap. To find a specific cohort definition id, refer to data/AllCohorts.csv which can be queried by disease name (cohortName), phenotyping algorithm (source), or cohort_definition id (cohortId)

Repository Structure

code/

Contains code used in repository, including:

• RunBasePhenotypesID.sh: query for 2+ occurrences of code
  • ROLLUP: Queries for descendants of code
  • CONTROL: Generates control cohort with 0 occurrences of code
• RunOHDSIPhenotypeLibraryID.sh: query for OHDSI phenotypes
• RunOMOPADOID.sh: query for ADO phenotypes
• RunOneOccID.sh: query for one occurrence of code
• RunPhecodeID.sh: query for Phecode phenotypes
• RunPhevaluatorDefns.sh: query for phenotypes used in PheValuator (extra specific and extra sensitive)
• test_phenotypes.py: testing suite

Subdirectories:

• cohort_json_files/: Contains json files or json file templates (in case where phenotyping algorithm follows the same template for all diseases)
• cohort_sql_files/: Contains sql files and sql file templates
• cohort_runs/: Example runs used to generate phenotypes
• PhenotypingScripts/: Helper scripts to render templates, translate codes to OMOP CDM and render sql