Trait Insights Gained by Comparing Genome-Wide Association Study Results using Different Chronic Obstructive...

Biobanks have facilitated the conduct of large-scale genomics studies, but they are challenged by the difficulty of validating some phenotypes, particularly for complex traits that represent heterogeneous groups ofpatients. The guideline definition of COPD, based on objective spirometry measures, has been preferred in genome-wide association studies (GWAS) conducted with epidemiological cohorts, but spirometry measures are seldom available for biobank participants. Defining COPD based on International Classification of Disease (ICD) codes or self-reported measures is highly feasible in biobanks, but it remains unclear whether the misclassification inherent in these definitions prevent the discovery of genetic variants that contribute to COPD. We found that while there was poor agreement in classification of UK Biobank participants as having COPD based on ICD diagnosis codes, self-reported doctor diagnosis or spirometry measures, contrasting GWAS results for these definitions provided insights into what patient characteristics each trait may capture.