analysis in bulk data (Supplementary Note and Supplementary Figs. 39,40). To correct for multiple testing, we confined the analysis to only test for primary cis- or trans-eQTLs that had a significant interaction with one or more cell types in MetaBrain Cortex-EUR (BH-FDR < 0.05), while also permuting the sample labels 100 times. Finally, we attempted replication of our findings using the snRNA-seq eQTL summary statistics of the recent preprint by Bryois and colleagues24. We overlapped their summary statistics with the Cortex-EUR cis-eQTLs and found (depending on the cell type) that between 9,402 and 13,764 overlapped. We calculated a BH-FDR on the P values of the ieQTLs that were significant in Cortex-EUR in the respective cell type. Given that the summary statistics did not include standard errors or MAF values, we predicted β and standard errors using the MetaBrain Cortex-EUR MAF together with the eQTL sample size, β and P value97 from Bryois et al. to calculate the Rb metrics.
