Integrative SMR and multi-model machine learning characterize FCER1A- and INSIG1-associated phenotypes in arterial aging-related disease

Background and aims : Global population aging has made age-related cardiovascular disease an urgent public health challenge, and inflammation is closely associated with arterial aging. This study aimed to identify inflammation-related molecular features associated with arterial aging-related disease using an integrative analytical framework. Methods : We integrated summary-data-based Mendelian randomization (SMR), machine learning, single-cell RNA sequencing, and weighted gene co-expression network analysis. SMR used FinnGen coronary artery disease GWAS and GTEx v8 whole-blood eQTL data. Models were evaluated in GEO transcriptomic datasets; single-cell and macrophage-state analyses resolved cell-type expression patterns. Expression was assessed in serially passaged human umbilical vein endothelial cells and naturally aged C57BL/6J mice. Drug-response prediction, docking, and molecular dynamics simulations prioritized candidate compounds and assessed modeled interactions. Results : FCER1A and INSIG1 were preferentially expressed in macrophage populations and varied across connected macrophage states and inferred trajectories in diseased vessels. Their expression was also increased during serial-passage-associated endothelial senescence and in aged mouse arterial tissue. Computational analyses prioritized dihydrorotenone and entospletinib and supported plausible modeled interactions with the corresponding protein structures. Conclusions : The integrative analysis prioritized FCER1A and INSIG1 as molecular features associated with arterial aging-related disease and provided cell-type localization and experimental expression evidence. These results set the stage for follow-up mechanistic and pharmacological investigations to dissect causal relationships and explore their therapeutic relevance.