Abstract
Overlap between primary sclerosing cholangitis (PSC) and systemic lupus erythematosus (SLE) has been documented in previous studies, but the precise causal relationships between them remain elusive. This study aims to clarify the potential causal links between PSC and SLE using Mendelian randomization (MR) and transcriptomic analyses. Genome-wide association study (GWAS) summary data for PSC and SLE were obtained from the IEU OpenGWAS database. The inverse variance weighted (IVW) method was the primary approach for assessing the causal links between PSC and SLE. Potential horizontal pleiotropy and heterogeneity were evaluated to ensure the reliability of the MR results. Additionally, transcriptomic analysis was performed using data from the Gene Expression Omnibus (GEO) database to uncover potential mechanisms insights involving overlapping genes and develop a diagnostic model for PSC and SLE. Genetically predicted PSC had a significant causal effect on SLE, with an OR of 1.190 (95% CI: 1.098-1.290, P ˂.001). Conversely, SLE had causal effect on PSC with an OR of 1.130 (95% CI: 1.046-1.221, P = .002). Multivariate MR analysis by adjustment of body mass index and smoking also revealed the bidirectional causal relationships between PSC and SLE. There were no signs of horizontal pleiotropy or heterogeneity, and the robustness of these results was confirmed via leave-one-out sensitivity analysis. Through integrated transcriptomic analysis and machine learning algorithms, 5 hub genes were identified. Furthermore, the diagnostic accuracy of the 5-gene signature was confirmed via Nomogram, calibration curve, and decision curve analysis, highlighting the diagnostic potential of these hub genes for both PSC and SLE. In conclusion, our MR study successfully confirms the bidirectional causal relationships between PSC and SLE, shedding light on the intertwined nature of these 2 conditions. Moreover, we identified 5 hub genes that potentially mediate the overlap between PSC and SLE, providing valuable insights for early diagnosis and future mechanistic exploration.