NEW YORK (GenomeWeb) – A pair of papers appearing in Nature Communications this week examined admixture among populations with African ancestry in the Western Hemisphere, and used that new dataset to study biases within genomic databases.
Researchers from the Consortium on Asthma among African-ancestry Populations in the Americas (CAAPA) deeply sequenced the genomes of nearly 650 individuals of African descent from the Americas and West Africa. They uncovered a continuum of admixture among the diaspora populations and refined estimates of deleterious mutations among people of African descent. In the second paper, researchers from the same consortium used that new dataset to examine how African ancestry affects genomic databases' determination of what might be a deleterious variant.
Both studies emphasized the need for better studies of populations of African and non-European ancestry, the researchers said.
"The African Diaspora in the Western Hemisphere represents one of the largest forced migrations in history and had a profound impact on genetic diversity in modern populations," Kathleen Barnes, the director of the Colorado Center for Personalized Medicine at University of Colorado School of Medicine and senior author of one paper, said in a statement. "Yet this group has been largely understudied."
In the first paper, CAAPA researchers led by Barnes deeply sequenced the genomes of 642 individuals of African descent living in the US, Colombia, Brazil, and Jamaica, among other Western Hemisphere countries, and in Nigeria. From this, they noted that individuals in their cohort harbored a greater number of single-nucleotide variants than global cohorts did, and that many of the SNVs were unique to their cohort.
This cohort represents populations with varying degrees of African ancestry, the researchers noted. For instance, they reported that Puerto Ricans in their sample had 27 percent African ancestry while Jamaicans had 89 percent African ancestry.
The patterns of rare genetic variation the researchers found among the various populations in their cohort largely reflected their population histories. For instance, among the Honduran Garifuna population, the researchers saw a lower number of mutations in only one individual and an increased median pairwise identity-by-descent, suggesting a series of bottlenecks. This population traces its roots, the researchers noted, to when a single slave ship from West Africa wrecked on a West Indian island in the 1600s.
Barnes and her colleagues also gauged the level of deleterious variants in their CAAPA cohort. On average, they estimated using PhyloPNH that each individual carried 1,625 deleterious SNVs. Individuals with a greater proportion of African ancestry carried more predicted deleterious heterozygotes, while individuals with a greater proportion of European ancestry carried more deleterious derived homozygotes — a difference the researchers ascribed to the original out-of-Africa migration.
"This [dataset] will contribute to the public database and give clinicians more information to better predict and track human disease," Barnes said. "It will allow us to tailor clinical [treatments] to specific individuals based on their ethnic and racial backgrounds."
In a companion paper, CAAPA investigators led by the University of Maryland School of Medicine's Timothy O'Connor applied this dataset to an evaluation of how commonly used genomic databases like ClinVar and Human Gene Mutation prioritize variants found among populations with African ancestry.
He and his colleagues classified variants as pathogenic annotated variants or non-annotated variants, based on their annotation in OMIM, HGMD, or ClinVar. They then sub-classified these variants as deleterious or non-deleterious based on computational predictions. From this, they noted a significant correlation between estimated African ancestry and the number of variants per individual in all variant sets, except deleterious pathogenic annotated variants.
In addition, when they removed those filters, they further noted a positive correlation between estimated African ancestry and variants in HGMD and in ClinVar.
O'Connor and his colleagues also noted that as ClinVar has updated its database, the association between African ancestry and pathogenic variants has changed. In particular, the researchers uncovered a reversal in this correlation between March and April of 2014, when a large dataset that included non-European individuals was added.
In addition, as more than 4.5 variants per person are likely to be prioritized for validation in people of African ancestry as compared to 2.8 variants per person in people of European ancestry, the researchers noted there's an added cost for reviewing genomic test results for people of African ancestry.
"[W]e need to expand these databases to include a broader range of ancestries, in order to produce more accurate medical genetic diagnoses," O'Connor said.