fig5

Figure 5. eQTL-based Mendelian randomization (MR) results. (A) Scatter plot showing SNP-level associations between COX4I1 expression (exposure) and localized scleroderma (LS) (outcome), with 95% confidence intervals (CIs). (B) MR effect estimates derived from the inverse variance weighted (IVW), MR-Egger, weighted median, weighted mode, and simple mode methods. The slopes of the colored lines represent the estimated causal effects. A downward trend indicates that higher COX4I1 expression is associated with a lower risk of LS. (C) Forest plot showing individual SNP-specific causal estimates calculated using the Wald ratio. (D) Leave-one-out sensitivity analysis evaluating the robustness of the MR estimates after sequential exclusion of each SNP. Statistical significance was primarily assessed using the inverse variance weighted (IVW) method, with P < 0.05 considered statistically significant. The MR-Egger, weighted median, weighted mode, and simple mode methods were used as complementary sensitivity analyses. MR: Mendelian randomization; eQTL: expression quantitative trait loci; LS: localized scleroderma; SNP: single nucleotide polymorphism; OR: odds ratio; CI: confidence interval.