Diverse UCLA Biobank Reveals New Genetic Links to Disease Risk and Drug Response

The report on the UCLA ATLAS Community Health Initiative describes analyses of linked genetic data and electronic health records that connected inherited variation with disease risk and medication response.

Investigators working with 93,936 participants reported multiple classes of findings, ranging from common-variant associations and polygenic risk score evaluations to pharmacogenomic signals and ancestry-enriched rare-variant links. The work is presented as a diverse, health-system–linked genetics resource intended to capture signals that can be difficult to observe in more homogeneous datasets.

The study reports that ATLAS participants span five continental ancestries and 36 fine-scale ancestry groups, with examples including Armenian, Ashkenazi Jewish, Filipino, and admixed American populations. Within the same health system, the authors describe this structure as enabling comparisons of genetic influences on health across populations while reducing differences in clinical care that can complicate across-system analyses. In that framing, the dataset’s mix of broad- and fine-scale ancestry labels is positioned as supporting discoveries that may not appear in less diverse resources, with diversity treated as shaping what signals can be detected.

Among pharmacogenomic observations, the reporting describes genetics as predicting how well patients respond to semaglutide used for weight loss, with treatment response varying across ancestry groups and being associated with genetic risk for type 2 diabetes. As a more specific example, the authors report that semaglutide response is associated with genetic variation in PTPRU. They characterize this semaglutide-linked signal as initial evidence supporting further investigation, keeping the description focused on association rather than mechanism. These results are presented as examples of how inherited variation tracked in ATLAS can align with observed differences in drug response.

The report also describes rare-variant analysis in ATLAS, noting that researchers examined rare genetic variants within the resource and highlighted ancestry-enriched signals emerging from that approach. Two examples are given: an association between ANKZF1 and peripheral vascular disease in individuals of African ancestry, and a relationship between EPG5 and HDL cholesterol and triglyceride levels in Ashkenazi Jewish individuals. The write-up frames these findings as gene–phenotype links identified within specific ancestry groups represented in ATLAS, without detailing effect sizes or biological pathways. In that context, the rare-variant results are presented as illustrations of ancestry-enriched discovery within the cohort. The study also reports additional gene–phenotype associations, including FN3K with intestinal disaccharidase deficiency in European and admixed American populations.

Polygenic risk score testing in ATLAS is described as showing that polygenic scores predicted common diseases, with effects diminished in non-European populations; however, the report notes that more work is needed to translate these findings into clinical practice across health systems.

Key Takeaways:

• The analysis is presented as a large, ancestrally diverse, health-system–linked biobank resource connecting genetics with EHR-derived phenotypes.
• Reported pharmacogenomic signals included ancestry-associated variation in semaglutide response and a semaglutide-associated signal involving PTPRU.
• Rare-variant and polygenic score evaluations highlighted ancestry-enriched associations and reported risk stratification patterns.