A new study by UCLA Health published in Cell on Mar. 27 reports that analyzing genetic data from a large, diverse biobank can reveal important connections between people's genes, disease risk, and how they respond to medications.
The research highlights the potential for personalized medicine by studying nearly 94,000 participants from the UCLA ATLAS Community Health Initiative Biobank. The findings suggest that including people from many different ancestries can uncover insights not seen in less diverse datasets. One notable result is that genetics may predict how well patients respond to GLP-1 drugs used for weight loss, with differences observed across ancestry groups and links to genetic risk for type 2 diabetes.
Dr. Daniel Geschwind, senior associate dean and associate vice chancellor of Precision Health at UCLA who oversees the ATLAS program, said: "This isn't a small lab finding. ATLAS represents a sweeping cross-section of real patients, making its discoveries directly translatable to the groups of people medicine has historically left behind." The study found new genetic associations such as a link between semaglutide response and the gene PTPRU.
Dr. Roni Haas, an assistant project scientist at UCLA Health and lead author of the paper, said: "Although many other efforts to integrate electronic health records with genetic data have advanced genetic and biomedical discovery, they've often had a heavy concentration of homogeneous populations of European ancestry, limiting generalizability." Haas added that Los Angeles County's diversity is reflected in ATLAS participants who span five continental ancestries and 36 fine-scale ancestry groups.
The research also examined rare genetic variants within specific ancestry groups. For example, it identified connections between ANKZF1 and peripheral vascular disease in African individuals and EPG5 with cholesterol levels in Ashkenazi Jewish individuals. The study evaluated polygenic risk scores—estimates based on multiple genes—to assess disease likelihoods among participants.
Geschwind said: "These findings showcase how UCLA Health's unique patient populations add significantly to understanding the genetic basis of medical disorders across the spectrum of disease and ancestry... we have pilot studies underway that we hope and expect will soon show the immediate clinical impact that this work has the potential to deliver." The publicly available web portal (https://atlas-phewas.mednet.ucla.edu) presents thousands of heritable associations across various diseases.
