Abstract
Despite advances in traditional risk factors for cardiovascular diseases (CVDs), significant residual risk of CVDs remains incompletely captured. Integrative analysis incorporating cardiovascular magnetic resonance imaging (CMR) could facilitate to discover novel -therapeutic targets. This study aimed to identify potentially druggable plasma proteins for CVDs by incorporating CMR traits with integrative omics analysis. By integrating protein quantitative trait loci (pQTL) datasets of plasma proteins from Atherosclerosis Risk in Communities (ARIC) study with genome-wide association studies of 19 CVDs and 82 CMR traits, we sequentially used proteome-wide association study (PWAS), Mendelian randomization (MR), and colocalization analysis to identify putatively causal proteins. Replication MR was conducted using deCODE pQTL data, followed by observational association analysis using UK Biobank individual-level data, and multidimensional downstream analyses, as well as phenome-wide MR (Phe-MR). In total, we identified 342 protein-CVD and 115 protein-CMR pairs through PWAS. MR and colocalization ana-lyses revealed 66 protein-CVD and 39 protein-CMR pairs with potential causal relationships, of which 51 protein-CVD and 33 protein-CMR pairs were replicated. Additionally, 26 protein-CVD and 6 protein-CMR pairs showed significantly observational associations. Multidimensional downstream analysis highlighted potential biological pathways and druggability insights. Notably, AGER, CCN3, FER, and SPON1 were identified as proteins associated with both CVDs and CMR traits. Phe-MR analysis suggests potential beneficial and adverse effects of these proteins on other diseases. Our findings highlight potentially druggable plasma proteins for CVDs by incorporating CMR traits, providing novel insights into CVD pathogenesis and therapeutic drug development.