A. Gaedigk and S. Leeder together with PharmGKB published an article in Genetics in Medicine about the prediction of CYP2D6 phenotypes from genotype data across world populations (http://www.nature.com/gim/journal/vaop/ncurrent/full/gim201680a.html).
In the supplemental materials of Clinical Pharmacogenetics Implementation Consortium (CPIC) guidelines involving CYP2D6 (https://www.pharmgkb.org/view/dosing-guidelines.do?source=CPIC and https://cpicpgx.org/guidelines/), allele frequencies across populations have been systematically captured and updated with each new guideline or update. Using this frequency table, the article describes the calculation of diplotype frequencies and the translation into phenotype based on the activity score system. The prevalence of genetically predicted poor, intermediate, normal, and ultrarapid metabolizer phenotypes are presented for major populations.
The article also includes a critical discussion highlighting the challenges of phenotype prediction from genotype data. This includes the overestimation of certain alleles in studies with minimal CYP2D6 genotyping and issues of grouping alleles and resulting diplotypes into functional and phenotype categories.