Challenges and disparities in the application of personalized genomic medicine to populations with African ancestry

Abstract

To characterize the extent and impact of ancestry-related biases in precision genomic medicine, we use 642 whole-genome sequences from the Consortium on Asthma among African-ancestry Populations in the Americas (CAAPA) project to evaluate typical filters and databases. We find significant correlations between estimated African ancestry proportions and the number of variants per individual in all variant classification sets but one. The source of these correlations is highlighted in more detail by looking at the interaction between filtering criteria and the ClinVar and Human Gene Mutation databases. ClinVar's correlation, representing African ancestry-related bias, has changed over time amidst monthly updates, with the most extreme switch happening between March and April of 2014 (r=0.733 to r=-0.683). We identify 68 SNPs as the major drivers of this change in correlation. As long as ancestry-related bias when using these clinical databases is minimally recognized, the genetics community will face challenges with implementation, interpretation and cost-effectiveness when treating minority populations.

Figure 1. Variant classification workflow and correlation with African ancestry.

(a) The pipeline used to categorize variation into four groups (deleterious (Del.) PAVs, non-deleterious (Non-del.) PAVs, deleterious NAVs and non-deleterious NAVs) each with different levels of clinical relevance (see Methods for further explanation). The three key filters used in separating variants are (1) MAF from multiple databases, (2) pathogenic annotation (as defined by the ClinVar and/or HGMD) and (3) deleterious prediction. For b–e the x axis is the proportion of African ancestry as estimated by ADMIXTURE. The corresponding y axes represent the total number of variants per individual for the following groups: (b) deleterious PAVs, (c) non-deleterious PAVs, (d) deleterious NAVs and (e) non-deleterious NAVs. Colours of each individual reflect the population sampling location.

Figure 2. Historical view of African-ancestry biases in ClinVar.

The x axis represents the various archived versions of ClinVar. In the black y axis on the right, we see the number of PAVs recorded from each version of the database. There are a few decreases in numbers, but overall this number shows continuous growth. In the blue y axis on the left, we see the correlation coefficient estimated between the number of PAVs per CAAPA individual and their proportion of African ancestry. The dotted grey line represents the date of the first official release of ClinVar. The blue trend line shows the instability across different ClinVar releases of the correlation of African-ancestry proportion with average number of pathogenic variants per individual. The change in correlation is particularly notable for sequential releases between March and April 2014, after which the correlation remains significantly negative for 3 months (April–July 2014) before once again becoming significantly positive. The red trend line represents the same relationship between ancestry–pathogenicity correlation and ClinVar release over time after applying filters, and shows a significant change in correlation during the same 3-month period of April–July 2014 despite an overall reduction in movement of the correlation across time.