Polygenic risk scores (PRS) were constructed using the results published PGC GWAS of SCZ (36,989 cases and 113,075 controls)32, BiP (20,352 cases and 31,358 controls)43, and MDD (135,458 cases and 344,901 controls)44. To obtain independent sets of SNPs, we performed linkage disequilibrium (LD) based “clumping” in the COGA EA sample (founders only) using a 500 kb physical distance and an LD threshold of r2 ≥ 0.1. For each COGA participant, we derived an individual-level risk score by weighting the number of risk alleles at each independent SNP by the natural logarithm of its odds ratio (OR) and summing this quantity across independent SNPs. scores based on nine progressively more inclusive thresholds of statistical significance in each discovery GWAS: p < 0.0001, 0.001, 0.01, 0.05, 0.1, 0.2, 0.3, 0.4, and 0.5. To aid in the interpretation, PRS values were transformed to Z-scores so that effect sizes could be expressed as SD of coherence values per SD of PRS.
