Why local ancestry matters
When we study human genetics, we rely heavily on reference databases to interpret DNA, guide diagnoses and inform medical research. But one of the first things we learn in this field is an uncomfortable truth: most of these datasets are built primarily from individuals of European ancestry.
This imbalance has real consequences. When genetic reference data does not reflect the diversity of global populations, it becomes harder to accurately interpret genetic variants in people from underrepresented groups. In practice, that means millions of individuals around the world may receive less accurate genetic information and fewer insights from genomic medicine.
In our recent study published in Nature Communications, our team set out to address part of this challenge. We developed and applied a pipeline that uses local ancestry inference (LAI) to analyze data from the Genome Aggregation Database (gnomAD), one of the largest publicly available genomic resources. Using this approach, we generated ancestry-specific allele frequencies for the inferred African/African American and Admixed American (Latino) genetic ancestry groups.
Traditional genomic analyses often average across the entire genome, thereby masking meaningful patterns of variation. LAI allows researchers to examine genomes at a much finer scale. Rather than treating an individual’s genome as belonging to a single ancestry group, LAI identifies the ancestral origin of each segment of DNA. This is especially important in admixed populations across the Americas, where many individuals have genetic ancestry from multiple continents due to historical migrations.
One striking finding from our work is the dramatic difference in allele frequency across ancestral segments within the same population. A genetic variant that appears rare when averaged across an admixed population may be common in one ancestral background and nearly absent in another. These differences matter because they can influence disease risk estimates and affect whether genetic tests correctly identify variants associated with disease. Allele frequencies are a foundational component of variant interpretation in both research and clinical genetics, so having improved estimates of frequencies can help rule out more variants as being potentially disease-causing. From a patient perspective, this has real consequences: it can reduce uncertainty, help avoid unnecessary follow-up testing and potentially prevent invasive interventions.
By generating ancestry-specific allele frequencies and making them available through gnomAD, our goal is to provide researchers, clinicians and genetic counselors with greater precision in interpreting genetic variation. Improving the accuracy of these resources is essential for ensuring that genomic medicine benefits individuals from all backgrounds.
By Pragati Kore, Ph.D. candidate in the Atkinson Lab at Baylor College of Medicine
