Frequency of pharmacogenomic variation and medication exposures among participants in the All of Us Research Program

Pharmacogenomics promises improved outcomes through individualized prescribing. However, the lack of diversity in pharmacogenomics studies impedes clinical translation and the application of precision medicine across populations. We evaluated the frequencies of variants in 16 pharmacogenes ( ABCG2 , CACNA1S , CYP2D6 , CYP2C Cluster , CYP2C19 , CYP2C9 , CYP3A5 , CYP4F2 , DPYD , G6PD , NUDT15 , RYR1 , SLCO1B1 , TPMT , UGT1A1 , and VKORC1 ) and predicted phenotypes and medication exposures using whole-genome sequencing and electronic health record data from nearly 100,000 All of Us Research Program participants. We report that 100% of participants carried at least one pharmacogenomic variant and nearly all (99.5%) had a predicted phenotype with prescribing recommendations. Clinical impact was high with 24.4% of participants having both an actionable phenotype for these pharmacogenes and a prior exposure to an impacted medication with pharmacogenomic prescribing guidance. Importantly, we report thousands of previously unrecognized pharmacogene variants and hundreds of alleles and predicted phenotypes that deviate from known frequencies across populations or were previously unreported, especially in the Americas and African ancestry groups.