Standardized Biogeographic Grouping System for Annotating Populations in Pharmacogenetic Research

Abstract

The varying frequencies of pharmacogenetic alleles among populations have important implications for the impact of these alleles in different populations. Current population grouping methods to communicate these patterns are insufficient as they are inconsistent and fail to reflect the global distribution of genetic variability. To facilitate and standardize the reporting of variability in pharmacogenetic allele frequencies, we present seven geographically defined groups: American, Central/South Asian, East Asian, European, Near Eastern, Oceanian, and Sub-Saharan African, and two admixed groups: African American/Afro-Caribbean and Latino. These nine groups are defined by global autosomal genetic structure and based on data from large-scale sequencing initiatives. We recognize that broadly grouping global populations is an oversimplification of human diversity and does not capture complex social and cultural identity. However, these groups meet a key need in pharmacogenetics research by enabling consistent communication of the scale of variability in global allele frequencies and are now used by Pharmacogenomics Knowledgebase (PharmGKB).

Figure 1:. Map of geographical boundaries included in each geographical population group.

Group boundaries for the seven geographical groups fall predominantly along national boundaries to aid the assignment of group membership. The two admixed groups of African American/Afro-Caribbean and Latino are not shown on this figure as the map indicates the borders of each geographical group based on the location of genetic ancestors pre-Diaspora and pre-colonization, which cannot be applied to the two admixed groups. It should also be recognized that, due to the large geographical areas covered by each group, a single group does not accurately represent the large amount of genetic diversity found in that one region.

Figure 2:. Principal component analysis comparing genetic distances of populations with close geographic proximity using 1000 Genomes and HGDP participants.

(A) The genetic gradient between populations is illustrated along PCs 1 vs 2 and PCs 3 vs 4, showing that, while completely discrete population boundaries are challenging, the groupings proposed here provide a statistically robust grouping. (B) AUCs of logistic regression to predict cluster membership, showing high degree of population structure. Note that, because none of the 1000 Genomes populations fall into the American (AME) group, no reference data were available to include this group in the analysis.

Figure 3:. Maps illustrating how the proposed biogeographical grouping system can be used to illustrate the variability in global frequencies of key pharmacogenetic alleles.

Allele frequencies from 1000 Genomes are shown across global populations for (A) CYP2D6*10, (B) CYP2C9*2, (C) CYP2C9*3 and (D) CYP2C9*8.