Abstract
Lung structures are critical for gas exchange and contribute to the pathogenesis of respiratory diseases, exhibiting notable lobe-specific heterogeneity. To investigate their genetic basis, we apply a deep-learning AI system and Pyradiomics to define lobe-specific lung CT imaging phenotypes, conducting genome-wide analyses in 35,469 participants from the Lung Imaging Genomics Initiative in China. We identify 36 loci associated with voxel intensities and 138 loci linked to three-dimensional shape. Functional annotation reveals significant enrichment of relevant genes in pathways regulating early fetal lung development and loci enriched in fetal lung regulatory elements. Genetic correlations are identified between lung structures and chronic respiratory diseases as well as lung function, with a number of loci showing colocalization. Mendelian randomization analyses suggest a causal role of lung structures in chronic lung diseases and extrapulmonary traits. This study provides new insights into the genetic architecture of lung structures and their links to diverse clinical outcomes.