Data and code for: Proteome-wide Mendelian randomization of circulating proteins in autism spectrum disorder and Tourette disorder

Analysis code, derived summary-level results, figures, and manuscript preprint supporting a proteome-wide two-sample Mendelian randomization study of 49 circulating proteins (UKB-PPP pQTL, n≈54,219) against autism spectrum disorder (ASD; ieu-a-1185, Grove et al. 2019) and Tourette disorder (TD; Strom et al. 2025). Lead signals were replicated with deCODE pQTL (SomaScan, n≈35,559) and evaluated with colocalization (coloc), heterogeneity, leave-one-out, MR-Egger, MR-PRESSO, and Steiger directionality. Among 49 proteins, no Bonferroni-significant causal signal was observed. The complement protein C2 showed a nominally significant MR signal (IVW p≈0.0065, risk direction) but colocalization favoured distinct causal variants (PP.H3 ≫ PP.H4), indicating the signal is driven by MHC-region linkage disequilibrium rather than a protein-mediated causal effect. C2 and EGF showed null MR against TD. Raw full-genome pQTL summary statistics are not redistributed (UKB-PPP ≈37 GB; deCODE ≈44 GB) and are obtained from the original sources; the deposited CSVs are the derived results produced by the included pipeline. License: CC-BY-4.0.