PheWAS

PheWAS performs phenome-wide association analyses that test associations between genetic variants and a broad spectrum of phenotypes using electronic medical records (EMRs) linked to genomic data.

Key Features:

• Integration with EMR data: Utilizes longitudinal EMR data linked to genetic information from biobanks such as BioVU and encodes phenotypes using International Classification of Disease (ICD9) billing codes.
• Code translation table: Employs a code translation table that automatically defines 776 distinct disease populations and matched controls by mapping prevalent ICD9 codes from EMR systems.
• Genotyping demonstration: Demonstrated by genotyping a cohort of European-Americans at five single nucleotide polymorphisms (SNPs) with known disease associations, including atrial fibrillation, Crohn's disease, carotid artery stenosis, coronary artery disease, multiple sclerosis, systemic lupus erythematosus, and rheumatoid arthritis.
• Replication of known associations: Replicated four of seven previously reported SNP–disease associations with statistically significant P-values ranging from 2.8 x 10^(-6) to 0.011.
• Discovery of novel associations: Identified 19 previously unreported statistical associations between the tested SNPs and various diseases at P < 0.01.
• Feasibility for large-scale studies: Demonstrates feasibility for large-scale SNP–disease association studies using existing EMR-linked biobank data.

Scientific Applications:

• Genetic epidemiology: Enables exploration of the genetic basis of diseases across diverse phenotypes in genetic epidemiology studies.
• Hypothesis-free screening: Supports hypothesis-free, phenome-wide screening to identify both known and novel genotype–phenotype associations.
• Application to EMR-linked datasets: Applicable to richly phenotyped datasets linked to EMRs and biobanks for broad-scale association discovery.

Methodology:

Translating ICD9 codes into defined disease populations, genotyping individuals at specific SNPs, and conducting statistical analyses to identify significant associations using EMR-linked genomic data.