Project Summary: Chronic obstructive pulmonary disease (COPD) is a debilitating disease for which new, disease-modifying treatments are desperately needed. Since drug targets supported by human genetic evidence are more likely to lead to FDA-approved treatments, functional characterization of genome-wide association study (GWAS) loci is a translational research priority. COPD GWAS led by our group and others have identified more than 80 significant loci. We have identified the target genes for some loci, including FAM13A, ACRV1B, and TGFB2, using integration of GWAS and expression quantitative trait loci (eQTLs) to guide functional studies. However, most COPD GWAS loci remain uncharacterized due to 1) lack of QTL maps in appropriate cell types and 2) genetic effects other than eQTLs, such as alternative splicing. Splicing QTL (sQTLs) have been shown to be as important as eQTL for functional characterization of GWAS loci, and we have used sQTLs from blood and lung to identify FBXO38 and NPNT as COPD GWAS splicing genes. In this proposal, we will use a new resource of 185 primary human airway epithelial cells (HAEC-185) cultured and exposed to cigarette smoke (CS) or Air with associated genome-wide genotyping and RNA-seq measurements. HAEC-185 data will be used to identify alternative splicing related to COPD and/or CS exposure in HAECs, a highly relevant COPD cell type. In Aim 1, we will generate the first HAEC genome- wide sQTL maps for fully differentiated HAECs exposed to CS or Air. We will perform multiple colocalization analysis between COPD GWAS and these sQTL maps to identify genes whose splicing is altered by COPD GWAS variants. For two loci, genetic effects on splicing will be confirmed via long read RNA sequencing and isoform-specific protein mass spectrometry in genotype-selected HAECs. In Aim 2, we will perform state-of-the-art fine mapping of COPD GWAS sQTLs that integrates deep learning splicing model variant predictions with Bayesian fine mapping methods. We will train new, cell-type specific deep learning splicing models using data from HAEC-185 and 1,376 lung tissue samples from the Lung Tissue Research Consortium. In Aim 3, we will extend our previous work in gene regulatory network modeling to develop new methods to identify splicing regulatory networks. These methods will be applied to CS exposure data in HAEC-185 to provide a holistic view of CS-associated alternative splicing and its impact on gene expression pathways and cellular phenotypes. The network analysis will identify key splicing regulators, i.e. RNA binding proteins (RBPs), whose effects will be studied using shRNA knockdown of selected RBPs in HAECs. Our multi-disciplinary research team has the requisite expertise in COPD genomics, airway epithelial biology, computer science, gene network inference, long read sequencing, and proteomics to complete this important project to characterize COPD and CS-related alternative splicing in HAECs.