News Release

New method developed to detect and adjust population structure in genetic summary data

Summix was developed by CU Denver researchers to increase equity in DNA databases, making them more useful for ancestries such as African-American and Latinx

Peer-Reviewed Publication

University of Colorado Denver

In a new study published today in the American Journal of Human Genetics, researchers announced the development of a new method to increase the utility and equity of large genetic databases. The research was conducted by Audrey Hendricks, an associate professor of statistics at the University of Colorado Denver (CU Denver).

Summix, the new method developed by Hendricks and her team of CU Denver undergraduate and graduate students, estimates the genetic ancestry in databases and adjusts the information to match the ancestry of a person or sample of people. This method leads large genetic databases to become more useful for people of various ancestries such as African American or Latinx, as they are underrepresented in genetic databases and studies. Hendricks compares this method to translating a book from English to another language.

"Think of DNA as the words of our body," says Hendricks. "All of the words of our body make the instruction book that makes each of us up. Right now, it's like the DNA books are only written in English so the information in the library is not as useful for people who don't speak English. We're working to create books in the library that are more universal."

According to Hendricks, individuals and samples from understudied populations, such as African American and Latnix, are the most likely to lack large public resources with precisely matched ancestry data. As a result, researchers working with those populations often resort to the closest, but still poorly matched ancestral group. This leads to biased results in the very populations where high-quality research is needed the most.

The team showed the effectiveness of Summix in over 5,000 simulation scenarios and in the widely used Genome Aggregation Database (gnomAD), a publicly available genetic resource. They found Summix's estimates of ancestry proportions to be highly accurate (within 0.001%) and the ancestry-adjusted genetic information to be less biased. The Summix method is available in open access software increasing the utility of the method and its applications.

"Most people are a combination of multiple continental (e.g. African and European) or finer scale (e.g. Italian and German) ancestries," said Hendricks. "As healthcare moves forward with precision medicine, matching the unique ancestral make-up of each person will become increasingly important. The ability of Summix to update a genetic resource to match the ancestry of an individual is an important step in this direction and helps to increase the utility and equity of genetic summary data."


This study was funded by the National Human Genome Research Institute through the Genome Sequencing Program and the 2020 Genomic Innovator Award.

About University of Colorado Denver

CU Denver is Denver's partner in progress and ally in innovation. Our connection to our vibrant city inspires leading research, creative work, and civic engagement. Our collaboration with Denver's businesses and local government helps set us apart from other universities. With a history that began in 1912, CU Denver has operated independently since 1973. Our location in downtown Denver serves more than 15,000 students. In Colorado and around the world, our talented graduates form a diverse and growing Lynx family. We work to create welcoming and respectful learning environments where a culture of inclusion can flourish. At CU Denver, we honor our diversity of experiences and perspectives in the committed belief that they enrich the educational experience for all. For more information, visit

Disclaimer: AAAS and EurekAlert! are not responsible for the accuracy of news releases posted to EurekAlert! by contributing institutions or for the use of any information through the EurekAlert system.