Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
Deep Learning Uncovers New Genetic Blueprint for Asthma Risk
A major study combining large-scale genetic analysis with advanced deep learning has identified hundreds of new genetic variants associated with asthma in Europeans, paving the way for improved risk prediction and personalized medicine approaches.
Background
Asthma affects millions worldwide, with genetic factors contributing significantly to disease susceptibility. While previous studies have identified risk variants, the complete genetic landscape—especially for European populations—remains incomplete. Researchers sought to leverage larger datasets and more sophisticated analytical methods to discover new asthma-risk genes.
Key Findings
This meta-analysis combined data from over 1.8 million individuals (158,763 cases). Standard genome-wide association analysis identified 69 previously unknown genetic loci. Incorporating eosinophil counts through multi-trait analysis revealed 32 additional novel loci, while conditional false discovery rate approaches identified 234 more. Most strikingly, a Transformer-based deep learning framework (InsightGWAS) prioritized 684 novel candidate variants. Key genes implicated include PIK3CD (immune signaling), AHR (environmental response), TNFRSF9 (T-cell activation), and NEGR1 (cell adhesion). Polygenic risk scores derived from these prioritized variants outperformed conventional approaches in predicting asthma liability.
Why It Matters
These findings establish a comprehensive genetic map for asthma in European ancestry populations and demonstrate that validated polygenic risk scores can effectively predict disease risk, supporting development of genetically informed risk stratification and precision medicine strategies.
Limitations
The study focused on European ancestry individuals; findings may not directly apply to other populations. Further validation in diverse cohorts strengthens generalizability claims.
Original paper: Deep learning and statistical methods identify novel asthma risk variants in Europeans. — The Journal of allergy and clinical immunology. 10.1016/j.jaci.2026.03.016
