Phenotype Risk Scores

Phenotype Risk Scores (PheRS) in Genomics England (GeL). In order to evaluate whether the PheRS can detect undiagnosed Mendlian disorders, this project calculates the PheRS for rare diseases in GeL and analyses the results between cases and controls.

Structure

folder	description
scripts	for analysis and creating datasets used
data	data and mappings used and/or created
phers	phers of all participants for diseases
phers/description	summary of codes mapped from each disease
samples	case/controls identified for diseases
samples/summary	summary of case/controls for diseases
diagrams	boxplot and histograms to compare case/controls

How to Use

Requires: GeL cohort datasets and ontology mappings in /data
Input: adjudicated GeL rare diseases mapped to OMIM, specificed in data/omimdb.csv
Run: /scripts/phers.R to calculate the PheRS for GeL rare disease cohort data in /data
Output:

• the phenotype risk scores for each disease in /phers
• descriptive information of the codes used for each disease in /phers/description
• case and matched control samples for each disease in /samples
• descriptive information about the case/control samples for each disease in /samples/summary
• visualisations of case/control phers distributions for each disease in /diagrams

Details

scripts

create_omimdb.R generates adjudicated dataset of GeL rare diseases mapped to OMIM IDs, /data/omimdb.csv

phers.R

1. Load mappings and gel data
2. Create functions to calculate PheRS:
   • create_filename
   • calculate_phers
   • create_case_control_sets
   • compare_case_controls
3. Run full analysis (functions above)

data

/data/omimdb.csv Normalised Specific Diseases | Phenotype Series | OMIM IDs