Integrative Analysis of eQTL Genes Reveals Key Biomarkers and Mechanisms for Early Diagnosis of Pancreatic Ductal Adenocarcinoma

This study presents a multi-omics, MR-informed framework for identifying eQTL-regulated biomarkers of PDAC. The identified hub genes offer promising avenues for early detection, while the mechanistic mapping of the PI3K-Akt pathway provides translational insights. These findings warrant further validation in clinical and experimental settings and hold potential to reshape PDAC diagnostic strategies.Pancreatic ductal adenocarcinoma (PDAC) remains a formidable clinical challenge due to its aggressive nature and lack of effective early diagnostic biomarkers. To address this, we integrated transcriptomic data, genome-wide association studies (GWAS), and expression quantitative trait loci (eQTL) information using Mendelian randomization (MR) to identify genes causally associated with PDAC risk. Differentially expressed genes were identified across 2 GEO datasets (GSE62165, GSE25471) and prioritized using weighted gene co-expression network analysis (WGCNA). MR analysis employing IVW, MR-Egger, weighted median, and MR-PRESSO identified 5 hub genes-CTSC, SMYD3, MFGE8, IGFBP7, and POC1B-as significant causal drivers of PDAC. These genes were incorporated into a diagnostic model constructed using machine learning approaches (random forest, SVM-RFE, Lasso), which achieved strong classification performance (AUC > 0.85) and excellent calibration (C-index = 0.92). Functional enrichment and protein-protein interaction analyses revealed that CTSC regulates the ECM-integrin-PI3K-Akt signaling pathway, contributing to tumor cell proliferation and survival. The findings establish a multi-omics-based biomarker panel with strong diagnostic utility and mechanistic relevance, suggesting a potential framework for future translational validation in clinical cohorts.