Asthma affects millions of people but does not look the same in every patient. Some develop asthma early in life, others later; some have mild disease, while others suffer frequent and severe flare-ups. These differences suggest that asthma is not a single condition, but a syndrome with different biological causes. Identifying these endotypes is essential to improve treatment and prediction.
Genes are known to influence asthma, but most studies look at one DNA change at a time. This approach has found many associations, yet it does not explain why patients differ so much in severity or in their response to medicines. Asthma is shaped by many genetic factors working together, and new approaches are needed to capture these complex patterns.
Our project unites genome-wide and machine learning approaches to uncover both individual genetic variants and broader genomic patterns that shape asthma heterogeneity. First, we will study genetic variants that predispose people to asthma and that influence response to standard medications. Second, we will use machine learning to group related DNA changes into modules, and then identify molecular subtypes of asthma by clustering patients according to these modules. Finally, we will test how these subtypes relate to severity, age of onset, obesity traits, and medication response.
By uncovering new genetic markers and clinically meaningful subtypes, this project will improve our ability to predict outcomes and guide more personalized treatments. The results may also create a framework for studying other common diseases where patients differ widely in risk and treatment response.
