NEW YORK (GenomeWeb News) – A study published online in PLoS ONE this week underscores the genetic variation that can exist amongst individuals sharing the same self-reported ancestry — findings that the authors say have implications for understanding and applying genetic information in a personalized medicine context.
Participants in the study were recruited through the Mount Sinai Biobank Project. That effort, based out of Mount Sinai Medical Center's Charles R. Bronfman Institute for Personalized Medicine (IPM) in New York, draws from the center's diverse, urban patient population.
The team genotyped nearly 1,000 individuals from the Biobank who self-identified themselves as European-American, African-American, or Hispanic. When they compared this genotype data with previously reported genetic information for other populations in the US, Brazil, and elsewhere, the researchers found a range of genetic ancestries and widespread heterogeneity across their self-reported patient populations.
"There's a continuum of mixtures of genetic ancestry in those individuals from African-American, European, or Native-American ancestry," IPM Director Erwin Bottinger, who was senior author on the study, told GenomeWeb Daily News. "That means that that population label — at least as far as the genetic risk and susceptibility is concerned — does not apply."
The IPM was launched roughly three-and-a-half years ago as part of Mount Sinai's strategy for finding ways to effectively translate genomic discoveries into clinical practice, Bottinger noted. Investigators at IPM want to assess the utility of published disease risk variants in this local population to determine which of these markers should be incorporated into clinical decision support algorithms.
"We need to begin to address the issue of how to translate genomic discoveries in a way that they are beneficial for all populations," he said. "We are an urban medical center where the patient population is very different from the patient populations that are typically studied in [genome-wide association studies], where it's case controlled and very homogeneous populations — overwhelmingly of European ancestry."
With such translational goals in mind, the IPM has recruited some 20,000 individuals for its Biobank Project, a repository containing DNA and blood samples from Mount Sinai patients who have agreed to provide these samples and to allow researchers access to their electronic medical records.
"The idea is that with this resource we have a way to begin to characterize the genetic background and the genetic risk profile of our local patient population, for the communities that we serve," Bottinger said.
For the current study, researchers used the Affymetrix 6.0 array to genotype 326 individuals who identified themselves as European American, 324 who said they were African American, and 327 who self-reported as Hispanic.
The team's subsequent ancestry and population structure analyses uncovered a range of genetic ancestry patterns within the patient populations.
For example, the Hispanic patient group contained sub-populations that clustered genetically with Mexican-American, Brazilian, and African-American populations assessed through previous studies such as HapMap.
But the researchers also found that Hispanic and African-American patients in the Biobank population formed a broader group with overlapping genetic patterns and varying degrees of genetic representation from different ancestral populations.
"These data forcefully underscore the diminishing relevance of the descriptors currently used for the two principle minority groups in the US," the study authors noted. "While consistent with previous descriptive studies, when viewed from the clinical perspective this evidence invites a reevaluation of the relevance of racial/ethnic labels."
In addition, when they began looking at how genetic variation in their patient populations related to associations detected through past genetic studies — focusing on seven SNPs across the FTO locus that have been linked to obesity in Europeans — the researchers found that all seven markers were associated with body mass index in the European-American group.
But associations between these SNPs and BMI was much more variable in the African-American and Hispanic patients, Bottinger explained, with certain SNPs corresponding to obesity in some of the African-American or Hispanic individuals but not others.
Results from the study highlight the need for caution when finding and characterizing genetic variants that may be relevant to clinical care for individuals from different populations. And, Bottinger says, they hint that individual genotyping is likely going to yield more relevant information for personalizing patient care than simply extrapolating risk and drug response information from studies of individuals from the same self-reported population.
"We will need to have a much more refined genetic approach, a much more detailed approach, to actually make sure that we, for an individual, capture the correct risk variant for clinical management," he explained.
The sorts of genetic variation present in a given patient population are also expected to vary from one location and medical center to the next, Bottinger said, noting that the Hispanic population at Mount Sinai, which is largely comprised of individuals from the Caribbean, likely shows different genetic profiles than Hispanic patient populations at medical centers in other parts of the country.
For their part, he and his team plan to genotype a set of disease-associated SNPs in the roughly 20,000 individuals enrolled in their Biobank Program in an effort to determine which associations replicate in this local population.
"That will give us a first real-life estimate and insight as to which markers can be applied across genetic backgrounds, which markers are independent of genetic background, which markers are strongly affected by genetic background, and therefore, can only be used in certain groups of individuals," Bottinger said.
"Based on our findings, we will move forward those markers where we understand very clearly their impact in various genetic backgrounds for disease risk," he said.