A new study by UCLA Health published in Cell presents a major advancement in the future of personalized medicine by pinpointing new connections between people’s genes, disease risk and medicine response by using a clinically well-characterized and diverse population-represented biobank.
By analyzing genetic data and electronic health records from 93,936 participants in the UCLA ATLAS Community Health Initiative Biobank , the research demonstrates how studying a broad range of ancestries within a single health system can reveal insights that may not emerge from less diverse datasets. This includes a striking new finding that genetics can predict how well patients respond to GLP-1 drugs for weight loss purposes. The analysis showed that treatment response to GLP-1 drugs varied across ancestry groups and was associated with a person’s genetic risk for type 2 diabetes. By integration with published proteomics, or sets of proteins produced by an organism or cell, from patients treated with these drugs, they were able to identify a genetic association of semaglutide response with the gene PTPRU. This finding provides the first evidence of a genetic link and supports further investigation of its activity. They were also able to replicate many previous studies in these data, highlighting UCLA Health's high-quality data.
“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,” said study senior author Dr. Daniel Geschwind , senior associate dean and associate vice chancellor of Precision Health at UCLA, who developed and oversees the ATLAS program.
The ATLAS Biobank reflects the wide range of ancestries living in Los Angeles, allowing researchers to compare genetic influences on health across populations while minimizing differences in clinical care that can complicate comparisons between health systems. This structure enables researchers to study how genetic variation contributes to disease and treatment outcomes in a real-world clinical setting.
“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,” said Dr. Roni Haas , an assistant project scientist at UCLA Health and the paper’s lead author. “UCLA Health serves Los Angeles County, one of the world’s most ancestrally diverse metropolitan areas, with a population of 9.6 million. The biobank reflects the diversity of Los Angeles: participants span five continental ancestries and 36 fine-scale ancestry groups, including Armenian, Ashkenazi Jews, Iranian Jewish, Filipino and Mexican American populations. This represents unusual diversity both within and outside European populations.”
The study also explored multiple layers of inherited variation, revealing previously unreported findings including differences in disease risk and healthcare use across ancestry groups, new genetic associations, and variation in how people respond to medications. For the first time in ATLAS, researchers analyzed rare genetic variants, discovering new links between genes and diseases within specific ancestry groups, including between the gene ANKZF1 and peripheral vascular disease in African individuals. In Ashkenazi Jewish individuals, a relationship was identified between EPG5 and HDL cholesterol and triglyceride levels. For ancestry groups enriched in ATLAS, new condition risks were identified. For example, Mexicans and South Americans showed increased susceptibility to adverse hormonal therapy effects.
The study also evaluated “polygenic risk scores,” which estimate a person’s likelihood of developing certain diseases based on many genetic variants across the genome. Testing polygenic risk score performance in ATLAS provided a very optimistic picture, suggesting that its clinical utility may not be far off for some traits, given the high fraction of diagnosed participants with elevated polygenic scores for diseases such as type 1 diabetes. Thousands of participants in ATLAS are estimated from the scores to be at high risk for many common diseases, but more work is needed to translate this into clinical practice across health systems.
In addition, the researchers used the diabetes and weight loss medication semaglutide, a GLP-1 RA (sold under brand names such as Ozempic and Wegovy), as a case study to examine how genetic variation may influence drug effectiveness. This found that genetics can predict how well patients respond to GLP-1 drugs for weight loss
“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,” Geschwind said. “Although this is one of the first concrete results at the research end of our efforts, 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.”
Studies that include diverse populations can help researchers discover genetic associations that may not appear in more limited datasets and may improve the usefulness of genetic tools across different populations. While most prior studies have used “broad-scale” genetic ancestry to link ancestry to health risk, the UCLA ATLAS study used broad- and fine-scale ancestries, which led to identification of numerous previously unreported genetic associations.
The ATLAS Community Health Initiative, launched in 2016 by the UCLA Institute of Precision Health to engage a diverse sample of patients across UCLA Health in precision health research, provides a clinical and genomic data resource for cutting-edge translational studies. Genetic information from this collection of diverse patient biospecimens is integrated with de-identified clinical information from the patient medical record and made available to UCLA-approved researchers. Today, more than 259,000 people have consented to participate, and more than 157,000 samples have been collected.
The publicly available web portal (https://atlas-phewas.mednet.ucla.edu) presents thousands of heritable genetic associations with diseases across diverse ancestry groups, along with disease risk estimates for more than 1,200 conditions.
The findings illustrate how combining genetic data with electronic health records in a large and diverse health system biobank can provide insights into both disease risk and treatment response, laying the groundwork for disease risk assessments and treatments tailored to each person's unique DNA.
Cell
Human tissue samples
Advancing Precision Health Discovery in a Genetically Diverse Health System
27-Mar-2026
E.E.K. has received personal fees from Regeneron Pharmaceuticals, 23&Me, Allelica, and Illumina; has received research funding from Allelica; and serves on the advisory boards for Encompass Biosciences, Overtone, and Galateo Bio. P.T.S. is a consultant for 10X Genomics, Illumina, Foresight Diagnostics, Natera, Exact Biosciences and Twinstrand. P.C.B. sits on the Scientific Advisory Board of Intersect Diagnostics Inc. and previously sat on those of Sage Bionetworks and BioSymetrics Inc. All other authors declare no conflict of interest.