RT Journal Article
SR Electronic
T1 Mixed-model admixture mapping identifies smoking-dependent loci of lung function in African Americans
JF bioRxiv
FD Cold Spring Harbor Laboratory
SP 626077
DO 10.1101/626077
A1 Andrey Ziyatdinov
A1 Margaret M. Parker
A1 Amaury Vaysse
A1 Terri H. Beaty
A1 Peter Kraft
A1 Michael H. Cho
A1 Hugues Aschard
YR 2019
UL http://biorxiv.org/content/early/2019/05/02/626077.abstract
AB Admixture mapping has led to the discovery of many genes associated with differential disease risk by ancestry, highlighting the importance of ancestry-based approaches to association studies. However, the potential of admixture mapping in deciphering the interplay between genes and environment exposures has been seldom explored. Here, we performed a genome-wide screening of local ancestry-smoking interactions for five spirometric lung function phenotypes in 3,300 African Americans from the COPDGene study. To account for population structure and outcome heterogeneity across exposure groups, we developed a multi-component linear mixed model for mapping gene-environment interactions, and empirically showed its robustness and increased power. When applied to the COPDGene study, our approach identified two 11p15.2-3 and 2q37 loci, exhibiting local ancestry-smoking interactions at genome-wide significant level, that would have been missed by standard singlenucleotide polymorphism analyses. These two loci harbor the PARVA and RAB17 genes previously recognized to be involved in smoking behavior. Overall, our study provides the first evidence for potential synergistic effects between African ancestry and smoking on pulmonary function and underlines the importance of ethnic diversity in genetic studies.